Osgoode Hall Law Journal Volume 59, Issue 2 (Spring 2022) Article 5 5-31-2022 The Dialogical Language of Law Julen Etxabe Peter A. Allard School of Law, University of British Columbia Follow this and additional works at: https://digitalcommons.osgoode.yorku.ca/ohlj Part of the Law Commons Article This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License. Citation Information Etxabe, Julen. "The Dialogical Language of Law." Osgoode Hall Law Journal 59.2 (2022) : 429-515. https://digitalcommons.osgoode.yorku.ca/ohlj/vol59/iss2/5 This Article is brought to you for free and open access by the Journals at Osgoode Digital Commons. It has been accepted for inclusion in Osgoode Hall Law Journal by an authorized editor of Osgoode Digital Commons. The Dialogical Language of Law Abstract We live in a dialogical world. The normative environment around us is many-voiced. Legal activities like drafting, negotiating, interpreting, judging, invoking, and protesting the law take place in dialogical encounters, all of which presuppose entrenched forms of social dialogue. And yet, the dominant modes of thinking about the law remain monological. How can we bring our legal conceptions into alignment with the dialogical world in which we live? The present article follows in the footsteps of a Bakhtinian dialogical theory of language that challenges the roots of contemporary positivist conceptions of law and language underpinning large swathes of legal academia and the legal profession—including recent approaches to legal interpretation called corpus linguistics. Against this backdrop, the article aims to develop a richer and more textured dialogical jurisprudence to encompass the various aspects, activities, and genres where legal language is employed. Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License. This article is available in Osgoode Hall Law Journal: https://digitalcommons.osgoode.yorku.ca/ohlj/vol59/iss2/5 429 The Dialogical Language of Law JULEN ETXABE* We live in a dialogical world. The normative environment around us is many-voiced. Legal activities like drafting, negotiating, interpreting, judging, invoking, and protesting the law take place in dialogical encounters, all of which presuppose entrenched forms of social dialogue. And yet, the dominant modes of thinking about the law remain monological. How can we bring our legal conceptions into alignment with the dialogical world in which we live? The present article follows in the footsteps of a Bakhtinian dialogical theory of language that challenges the roots of contemporary positivist conceptions of law and language underpinning large swathes of legal academia and the legal profession—including recent approaches to legal interpretation called corpus linguistics. Against this backdrop, the article aims to develop a richer and more textured dialogical jurisprudence to encompass the various aspects, activities, and genres where legal language is employed. * Assistant Professor and Canada Research Chair in Jurisprudence and Human Rights (Tier II), Peter A Allard School of Law, UBC. I wish to thank James Boyd White, Peter Goodrich, Hoi Kong, Gary Watt, Mónica López Lerma, Jeremy Webber, and the anonymous reviewers of Osgoode Hall Law Journal for their thoughtful and generous readings. I also want to thank Mark Antaki and the rest of participants in the research group Humanités Juridiques for inviting me to present an earlier draft at McGill University in March 2020. 430 (2022) 59 OSGOODE HALL LAW JOURNAL 1 ................................................................................................................................................... 437 I. LANGUAGE AS LIVING SPEECH ........................................................................................................ 437 II. THE UTTERANCE VERSUS THE SENTENCE ..................................................................................... 450 III. DIALOGICAL RELATIONS AND THE INTERNAL DIALOGISM OF DISCOURSE .................................. 462 2 ................................................................................................................................................... 471 IV. THE LANGUAGE OF RULES AND LEGAL POSITIVISM ...................................................................... 471 V. CORPUS LINGUISTICS AND STATUTORY INTERPRETATION ........................................................... 473 VI. THE ALLEGED NEUTRALITY OF DESCRIPTIVE LEGAL POSITIVISM ................................................ 483 VII. DEBUNKING “DETACHED NORMATIVE STATEMENTS” ................................................................... 490 VIII. THE GRAMMAR OF LAW AND THE GAME OF CHESS ....................................................................... 501 IX. CONCLUSIONS .................................................................................................................................. 511 “Language is not a neutral medium that passes freely into the private property of the speaker’s intentions; it is populated—overpopulated—with the intentions of others. Expropriating it, forcing it to submit to one’s own intentions and accents, is a difficult and complicated process.”1 “Most of what we say…comes to us already in words, already languaged.”2 WE LIVE IN A DIALOGICAL WORLD. We all exist in a whirlwind of exchanges, resonances, reverberations, and echoes. The normative environment around us is many-voiced, with multiple actors speaking in different registers and accents. Legal activities like drafting, negotiating, interpreting, judging, invoking, even protesting the law, take place in dialogical encounters. The basic “genres” of law— legislation, adjudication, administration, doctrinal elaboration—presuppose deep forms of social dialogue. From above, transnational processes of judicial dialogue, cross-referencing, and mutual citation expand the circulation of legal materials that disrupt the traditional closure of legal sources;3 from below, or in parallel, internal pluralism and non-state forms of law challenge the state’s presumptive 1. 2. 3. Mikhail M Bakhtin, “Discourse in the Novel” in Michael Holquist, ed, The Dialogic Imagination (University of Texas Press, 1981) 259 at 294 [Bakhtin, “Discourse”]. Alton L Becker, Beyond Translation: Essays Toward A Modern Philology (University of Michigan Press, 1995) at 287. Ann-Marie Slaughter, A New World Order (Princeton University Press, 2005). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 431 monopoly over law making.4 And yet, the dominant positivist modes of thinking, practicing, and teaching law remain monological. How can we bring our legal conceptions into alignment with the dialogical world in which we live? In what language can we even proceed? The present article follows the footsteps of Mikhail Bakhtin, who developed an entirely novel dialogical theory of language. His theory of language as dialogic presents a radical challenge to the dominant conception of language as an arbitrary system of signs, most famously developed by Ferdinand de Saussure.5 Beneath the poignant critique of Saussurian linguistics, however, there lies a deeper philosophical objection to positivist and propositional conceptions of language labelled “abstract objectivism,” the intellectual roots of which go back to Descartes and the rationalism of the seventeenth and eighteenth centuries (if not even earlier to neo-Platonism and medieval scholasticism). These roots have survived almost unscathed to this day—most recently, in Artificial Intelligence, algorithmic decision making, and the approach to statutory interpretation known as corpus linguistics. Against this backdrop, Bakhtin presented a conception of language as dialogical, interactive, and essentially evaluative. This article aims to flesh out this theory in the particular context of law. While reconstructing the views of a towering figure in twentieth-century philosophy, literary theory, and phenomenology is significant in itself, my underlying goal is to bring it to bear on contemporary jurisprudence, where 4. 5. On the significant work done by Indigenous scholars, see Christine Black, The Land is the Source of the Law: A Dialogic Encounter with Indigenous Jurisprudence (Routledge, 2011); Irene Watson, Aboriginal Peoples, Colonialism and International Law (Routledge, 2015); John Borrows, Freedom and Indigenous Constitutionalism (University of Toronto Press, 2016); Gordon Christie, Canadian Law and Indigenous Self-Determination: A Naturalist Analysis (University of Toronto Press, 2019). On the internal pluralism of law, see Kirsten Anker, Declarations of Interdependence: A Legal Pluralist Approach to Indigenous Rights (Ashgate, 2014). The most poignant exposition of this critique can be found in Bakhtin’s close collaborator Valentin Voloshinov’s book, which has been sometimes attributed to Bakhtin (see infra notes 25, 26). See Marxism and the Philosophy of Language, translated by Ladislav Matejka & I R Titunik (Harvard University Press, 1973) [Voloshinov, Philosophy of Language]. 432 (2022) 59 OSGOODE HALL LAW JOURNAL it has received little detailed attention.6 As I will show, Bakhtin’s criticism has its greatest bite against a reified image of law as a system of rules, which is at the heart of modern legal positivism and continues to underpin large swathes of legal academia, the legal profession, and the public at large.7 The abstract philosophical style targeted by Bakhtin still undergirds contemporary analytical jurisprudence, namely the “Oxbridge style” of philosophizing that still dominates Anglo-American, European, and Latin American schools8—in spite of some avowed support for the more pragmatic and ordinary-language views of J.L. Austin, Ludwig Wittgenstein, and others.9 Whether or not legal positivists of different persuasions consciously rely on the view of language criticized by Bakhtin, they purport to capture conceptual truths about the law by means of 6. 7. 8. 9. According to Desmond Manderson, the discussion of Bakhtin has been thin and often simply gestural, even in a field of study as attentive to language as law and literature. See “Mikhail Bakhtin and the Field of Law and Literature” (2016) 12 L Culture & Humanities 221 at 222-23. One early and significant exception is Peter Goodrich, who can be considered a precursor of sociolinguistic and discourse analysis. See Legal Discourse: Studies in Linguistic, Rhetoric and Legal Analysis (MacMillan, 1987) [Goodrich, Legal Discourse]. For a socio-legal approach to Bakhtin, see Mariana Valverde, Chronotopes of Law: Jurisdiction, Scale and Governance (Routledge, 2015). For other attempts to bring Bakhtinian ideas to law, see e.g. Charles Hersch, “Bakhtin and Dialogic Constitutional Interpretation” (1994) 18 Leg Stud F 33; Robert Rubinson, “The Polyphonic Courtroom: Expanding the Possibilities of Judicial Discourse” (1996) 101 Dick L Rev 3. Legal positivism is a longstanding tradition starting with Bentham and Austin that gained momentum in the twentieth century with the major works of Hans Kelsen and HLA Hart. Though not exactly identical, for the most part I will use legal positivism and analytical jurisprudence interchangeably. To be sure, it is possible to be a legal positivist and not an analytical philosopher, and vice versa. For important exponents of the Italian school, see Anna Pintore & Mario Jori, eds, Law and Language: The Italian Analytical School, translated by Zenon Bankowski, Simona Stirling & Anna Pirrie (Deborah Charles, 1997). In the Spanish-speaking world, see Manuel Atienza & Juan Ruiz Manero, A Theory of Legal Sentences (Kluwer Academic, 1998). See Brian Bix, Law, Language, and Legal Determinacy (Oxford University Press, 1993). Hart himself appeared to find important lessons in JL Austin, though it is unclear whether he internalized them (see infra note 176). Joseph Raz too appears to have borrowed from Austin, though with questionable effect (see infra note 328). I do not wish to engage in a sustained argument about ordinary-language philosophy, although in one important sense, the path of analytical philosophy runs contrary to ordinary-language philosophers such as Gilbert Ryle, Wittgenstein, Paul Grice, and Austin, who did not mean to simplify philosophical problems by bringing them down to ordinary language, but rather intended to show that ordinary language was already fecund ground for philosophical investigation. At any rate, legal positivists who attend to ordinary language (e.g., the speech-act theory of John Searle) emphasize the rule-governed character of language and hence the qualities of language that are system-like (see infra note 175 and section VIII, below). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 433 discrete propositions (or theses) that can be stated and verified in a true-or-false fashion as a matter of fact.10 In this vein, the kind of linguistic analysis performed by legal positivism and analytic jurisprudence falls squarely within the patterns of abstract objectivism that Bakhtin criticized. Legal positivists may disagree about the particulars of their commonalities, but their debates tend to center around the ideas that law can be identified by social facts alone (the social sources thesis); that law is not necessarily connected to morality (the separation thesis); and that the task of jurisprudence is generally descriptive, requiring no personal evaluation (the methodological thesis).11 Not all positivists agree with all of these, and in fact part of what defines the positivist camp is the debate about what constitutes the “necessary and sufficient” conditions of law.12 It is not my purpose to confront these theses directly13 but rather to challenge the linguistic assumptions upon which they are built, including the idea that law can be explained, articulated, or contained in any thesis of this kind. That law could be reduced to a set of discrete propositions is, to my mind, an impoverished understanding that has detrimental consequences for the kind of law that is being taught, studied, and imagined.14 Beyond any specific views, what actually unites the positivist camp is the assumption that law operates in society as a system of norms that can be described in propositional terms, independently from the theorist’s evaluative perspective.15 This widespread notion—shared also by “new realists” and socio-legal scholars 10. See Julie Dickson, Evaluation and Legal Theory (Hart, 2001) [Dickson, Evaluation]. See also Andrei Marmor, Philosophy of Law (Princeton University Press, 2011). 11. Andrei Marmor mentions another core positivist commitment, namely, the idea that law is essentially a means or instrument that can be put to good or bad uses. See “Legal Positivism: Still Descriptive and Morally Neutral” [Marmor, “Still Descriptive”] in Law in the Age of Pluralism (Oxford University Press, 2007) 125 at 128. For a general critique of the instrumental view of law, see Brian Tamanaha, Law as a Means to an End: Threat to the Rule of Law (Cambridge University Press, 2006). 12. Jules L Coleman has argued somewhat contentiously that the separability thesis (staunchly defended by Hart) is not central to legal positivism. See “Legal Directives and Moral Reasons” (Lecture delivered at Princeton University, 6 November 2008), online (pdf ): <lapa. princeton.edu/sites/default/files/COLEMANLegal_Directives_and_Moral_Reasons.pdf>. Likewise, John Gardner argued that the positivists only agree on the sources thesis, whereby the criteria to decide whether a norm is legally valid depend on its sources, not its merits. See Law as a Leap of Faith (Oxford University Press, 2012) at 21. 13. For a thorough critique, see Margaret Martin, Judging Positivism (Hart, 2014). 14. For a similar argument addressed to the field of analytical jurisprudence, see Costas Douzinas & Adam Gearey, Critical Jurisprudence: The Political Philosophy of Justice (Hart, 2005). 15. See Dickson, Evaluation, supra note 10. 434 (2022) 59 OSGOODE HALL LAW JOURNAL who do not necessarily espouse a positivist conception of law16—presupposes and relies upon an abstract, impersonal, and instrumental notion of language.17 Because the positivist view of language and the positivist conception of law are mutually reinforcing, a careful exposition of the Bakhtinian dialogical conception of language serves to make positivist arguments about law lose ground. Beyond the reconstructive and the critical, this article has a third important aim, which is to weave the threads of a dialogical jurisprudence that can encompass all aspects, activities, and genres that employ legal language. I will show how the positivist conception of language common to legal doctrine, legislation, statutory interpretation, and adjudication, can be replaced with a richer, more textured dialogical conception. Rather than construct this theory from the ground up, I pursue my argument through a close-up criticism of legal positivist assumptions about language. This might call for a bit of explanation. I argue that positivism remains the default position in the social imaginary of legal academics, practitioners, and the public at large, including those who approach law from neighbouring disciplines (sociology, political science, economics, et cetera). At the same time, legal positivism has been subject to powerful and varied forms of criticism from American realists, critical legal studies scholars, Dworkinians, legal pragmatists, and sociolegal scholars, as well as many contemporary feminists, critical race theorists, and post-colonial theorists. The question is: What does Bakhtin add that has not already been said multiple times? Many of the critics listed above have convincingly argued that legal positivism fails to live up to its promise of neutrality and objectivity; that it is unable to separate law from morality (and/or politics); that it is ideologically fraught with its own mythologies; and that it is enmeshed in state violence and the hegemonic 16. On the assumptions of the “new realists” see Julen Etxabe, “Law as Politics: Four Relations” (2020) 16 L Culture & Humanities 24. For the argument that socio-legal scholars “forget” about language, see Marianne Constable, “Genealogy and Jurisprudence: Nietzsche, Nihilism, and the Social Scientification of Law” (1994) 19 Law & Soc Inquiry 551; Marianne Constable, “Thinking Nonsociologically about Sociological Law” (1994) 19 Law & Soc Inquiry 625. 17. See e.g. Andrei Marmor, The Language of Law (Oxford University Press, 2014) [Marmor, Language of Law]. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 435 discourses of power and hierarchy.18 My target, however, is the very underlying conception of language that gets positivism off the ground. In this regard, a closer engagement with the linguistic presuppositions of legal positivism reveals that some of these assumptions are in fact shared by critics, who emphasize the instrumental quality of legal language, approach legal-linguistic competence as a form of cognitive mastery, or seek conceptually to determine the deep structure or grammar of legal language. Bakhtin is useful to avoid reproducing the same reified notion of language that underpins positivism. By following the trace marked by legal positivism, this article seeks a complete redescription of law with very different dialogical postulates. In other words, it is not that I wish to elaborate a dialogical conception of language and then apply it to law. Rather, my inquiry into a Bakhtinian notion of language is, at the same time, a dialogical theory of law. Finally, some of the best work presently done in legal theory (on space, movement, materiality, affect, sensory experience, visuality, and performance)19 appears to signal a desire to leave behind, or at least displace, the primacy of the textual and linguistic basis of law. From this perspective, a return to Bakhtin should not be interpreted as an attempt to reinscribe an outdated version of language. Bakhtin approached language as social, material, multi-sensorial, and 18. For powerful critiques of positivism, see Peter Fitzpatrick, Sociology of Law and Crime: The Mythology of Modern Law (Routledge, 1992); Scott Veitch, Law and Irresponsibility: On the Legitimation of Human Suffering (Routledge-Cavendish, 2007); Marianne Constable, Just Silences: The Limits and Possibilities of Modern Law (Princeton University Press, 2005). See also Panu Minkkinen, Thinking Without Desire: A First Philosophy of Law (Hart, 1999). For a deconstructive and feminist perspective, see Drucilla Cornell, The Philosophy of the Limit (Routledge, 1992). 19. On space, see Andreas Philippopoulos-Mihalopoulos, Spatial Justice: Body, Lawscape, Atmosphere (Routledge, 2015). On the senses, see Sheryl N Hamilton et al, eds, Sensing Law (Routledge, 2017). On sound, see James EK Parker, Acoustic Jurisprudence: Listening to the Trial of Simon Bikindi (Oxford University Press, 2015). On movement, see Olivia Barr, A Jurisprudence of Movement: Common Law, Walking, Unsettling Place (Routledge, 2016). On materiality, see Cornelia Vismann, Files: Law and Media Technology, translated by Geoffrey Winthrop-Young (Stanford University Press, 2008). See also Hyo Yoon Kang & Sara Kendall, “Contents, Introduction & Contributors” (2019) 23 Text Culture 1. On performance, see Marett Leiboff, Towards a Theatrical Jurisprudence: Space, Materiality, and the Normative (Routledge, 2020). For an excellent compendium, see Margaret Davies, Law Unlimited: Materialism, Pluralism, and Legal Theory (Routledge, 2017). 436 (2022) 59 OSGOODE HALL LAW JOURNAL aesthetic all at once.20 And while there is no extra-linguistic reality that can be directly accessed, it is possible, and indeed worthwhile, to tap into a conception of language that emphasizes the historically situated and uninterrupted stream of social interaction in which language comes alive.21 The article is divided into two halves. The first half of the article (1. Sections I-III, below) outlines a Bakhtinian view of language. Because everything that follows in the second half flows from this, I have found it necessary to elaborate on Bakhtin’s theory of language in some detail in order to carefully tease out some of its key legal implications. Section I begins with a general overview of language as social, interactive, and living, in contrast to the Saussurean notion of language as a system. Section II contrasts utterances as described by Bakhtin with sentences as relied on by legal positivists, before moving to the key dialogical nature of language (Section III). The second half of the article (2. Sections IV-VIII, below) pits this Bakhtinian view of language directly against discrete problems of legal positivist theory: the emphasis on rules and propositions (Section IV); the new methodology of corpus linguistics for statutory interpretation (Section V); the descriptive project of legal positivism (Section VI) and so-called detached normative statements (Section VII); and, finally, the idea of the system to bring unity to the whole, as exemplified in the metaphor of chess (Section VIII). Throughout I will argue that the view of language relied upon by legal positivists renders them unable to seize their 20. Writing in 1984 and again in 1987, Peter Goodrich was able to correctly pinpoint that despite the central importance of language for legal theory and legal practice, there was virtually no coherent or systematic account of the relation of law to language. Even worse, he considered, the occasional incursions of modern jurisprudence into the field had been exercises aimed at asserting or defending the positivistic view of law. See Peter Goodrich, “Law and Language: An Historical and Critical Introduction” (1984) 11 JL & Soc’y 173 at 181 [Goodrich, “Law and Language”]; Goodrich, Legal Discourse, supra note 6 at 1. The voluminous literature on law and language since Goodrich’s diagnosis may require qualifying his statement, but as this article seeks to demonstrate, it cannot entirely dispel it. On the language of law, see Peter M Tiersma & Lawrence M Solan, eds, The Oxford Handbook of Law and Language (Oxford University Press, 2012); Michael Freeman & Fiona Smith, eds, Law and Language, vol 15 (Oxford University Press, 2013). On language as a material, embodied, and gestural—i.e., “multimodal”—practice, see Gregory Matoesian & Kristin Enola Gilbert, Multimodal Conduct in the Law: Language, Gesture and Materiality in Legal Interaction (Cambridge University Press, 2018). For a concise and helpful introduction to the various trends in law and language scholarship, see Penelope Pether, “Language” in Austin Sarat, Matthew Anderson & Catherine O Frank, eds, Law and the Humanities: An Introduction (Cambridge University Press, 2010) 315. 21. For a Heideggerian approach noting the musicality of language, see Mark Antaki, “No Foundations?” (2014) 11 No Foundations: An Interdisciplinary J L & Justice 61. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 437 object of study in the neutral and detached fashion they assert, nor even when advocated as an ethical or normative ideal (Section IX). In conclusion, I will pave the way for a dialogical theory of law and judgment. 1 I. LANGUAGE AS LIVING SPEECH “The actual reality of language is not the abstract system of language forms…but the social event of communication.”22 Aside from his credentials as a literary critic, social theorist, and moral philosopher, Mikhail Bakhtin made seminal contributions to the philosophy of language.23 The main tenets of his views were developed to different degrees of specificity and variation in several important books and articles, including his 1929 monograph on Dostoevsky, the long essay “Discourse in the Novel” (1934–35), and other important articles on speech genres (1952–53) and other aspects of utterances, 22. Voloshinov, Philosophy of Language, supra note 5 at 94 [translation slightly modified]. 23. The prestige of Bakhtin as a literary theorist is underscored by his monographs on Dostoevsky. See Problems of Dostoevsky’s Poetics, ed & translated by Caryl Emerson, vol 8 (University of Minnesota Press, 1984) [Bakhtin, Dostoevsky]; Mikhail Bakhtin, Rabelais and His World, translated by Helene Iswolsky (Indiana University Press, 1984) [Bakhtin, Rabelais]; Mikhail Bakhtin, “The Bildungsroman and Its Significance in the History of Realism (Toward a Historical Typology of the Novel)” in Caryl Emerson & Michael Holquist, eds, Speech Genres and Other Late Essays, translated by Vern W McGee (University of Texas Press, 1986) 10 [Bakhtin, Late Essays]. On Bakhtin as social theorist, see Michael Gardiner, The Dialogics of Critique: M.M. Bakhtin and The Theory of Ideology (Routledge, 1992). See also Ken Hirschkop, Mikhail Bakhtin: An Aesthetic for Democracy (Oxford University Press, 1999). On Bakhtin’s early phenomenological works, see Augusto Ponzio & Susan Petrilli, Philosophy of Language, Art and Answerability in Mikhail Bakhtin (Legas, 2000). Of the extensive bibliography on Bakhtin, see especially Tzvetan Todorov, Mikhail Bakhtin: The Dialogical Principle, translated by Wlad Godzich, vol 13 (University of Minnesota Press, 1984) [Todorov, The Dialogical Principle]; Craig Brandist, The Bakhtin Circle: Philosophy, Culture, and Politics (Pluto Press, 2002) [Brandist, The Bakhtin Circle]; Michael Holquist, Dialogism: Bakhtin and his World, 2nd ed (Routledge, 2002); Frank Farmer, ed, Landmark Essays on Bakhtin, Rhetoric, and Writing, vol 13 (Routledge, 1998). For an excellent brief introduction, see Alastair Renfrew, Mikhail Bakhtin (Routledge, 2015). 438 (2022) 59 OSGOODE HALL LAW JOURNAL live communication, intertextuality, and discourse.24 Two additional works that are key to grasping Bakhtin’s views on language are Marxism and the Philosophy of Language (1929) and “Discourse in Life and Discourse in Art” (1926), both published by Valentin Voloshinov, most likely in close collaboration with Bakhtin. While Bakhtin himself was reluctant to confirm or deny his hand in the writings that were belatedly attributed to him, it is undeniable that these works share a “common conception of language” sustained throughout Bakhtin’s life, despite natural changes in focus and terminology.25 For the purpose of reconstructing Bakhtin’s view of language here, I will treat these later works as attributable to Bakhtin—i.e., authorized, though not necessarily authored, by him26—while my formal citations reflect the names of the originally published authors. In order fully to appreciate Bakhtin’s radical challenge, it is helpful to understand the prevalent view he resisted: The model of language as a system.27 24. See especially Mikhail M Bakhtin, “The Problem of Speech Genres” in Bakhtin, Late Essays, supra note 23, 60 [Bakhtin, “Speech Genres”]; Mikhail M Bakhtin, “The Problem of the Text in Linguistics, Philology, and the Human Sciences: An Experiment in Philosophical Analysis” in Bakhtin, Late Essays, supra note 23, 103 [Bakhtin, “Text”]. 25. The problem of authorship is a vexed one since the Soviet linguist Vyacheslav Ivanov raised the idea that Bakhtin was behind these writings. Bakhtin spoke reluctantly about the subject, but when challenged acknowledged that he wrote three books for his friends: Voloshinov (Freudianism (1927)) and Philosophy of Language (1929)) and Medvedev (The Formal Method in Literary Scholarship (1928)). See Sergey Bocharov, “Conversations with Bakhtin” (1994) 109 PMLA 1009 at 1013-14. In a letter to V.V. Kozhinov, dated 10 January 1961, Bakhtin wrote that these books were written in the “closest creative contact” and are based on a “common conception of language,” which nevertheless does “not diminish the independence and originality of each” (ibid at 1016). Bakhtin continued: “To this day I hold to the conception of language and speech that was first set forth, incompletely and not always intelligibly, in these books, although the concept has of course evolved in the past thirty years” (ibid). 26. Even considering that Bakhtin wrote the core ideas of the books, attributing every statement to him is not a straightforward issue, because Bakhtin admitted that he did not write these books as he would have had they been published under his own name. See ibid at 1015, 1017. Moreover, we cannot exclude that Voloshinov or Medevedev finished or adapted the texts along the patterns of this “common conception of language” (ibid at 1017). 27. In typical dialectical fashion, Marxism and the Philosophy of Language situates its own conception of language between two antagonistic trends: abstract objectivism and individualistic (or expressive) subjectivism. Whereas Saussure is the leading figure in the objective trend, subjectivism finds its roots in Romanticism and the work of Wilhelm von Humboldt (with Hamann and Herder as predecessors). See Wilhelm von Humboldt, ‘On Language’: On the Diversity of Human Language Construction and its Influence on the Mental Development of Human Species, ed by Michael Losonsky, translated by Peter Heath (Cambridge University Press, 1999) at 49. For Humboldt, language is activity and unceasing process of creation (energeia) realized in individual speech acts, rather than ready-made product (ergon). In this subjectivist trend, the laws of language creativity are the laws of individual psychology. Some of the best well-known linguists in Bakhtin’s time were Karl ETXABE, THE DIALOGICAL LANGUAGE OF LAW 439 Saussure’s point of departure was the tripartite distinction between langage (the human capacity for speech); langue (a given language system, e.g., Latin or French), and parole (concrete instances of uses of language, i.e., individual sentences).28 According to Saussure’s formulation, langue and parole belong to the natural reality of langage, a many-sided phenomenon combining physical, physiological, and psychological aspects deemed too heterogeneous to study under unitary principles. In turn, parole is purely individual and depends upon the will and intellect of speakers, once again deemed too contingent for systematic study. Given that neither langage nor parole may be the object of scientific study, Saussure argued that the linguist has to focus on langue, namely a homogeneous, bounded, and closed system of arbitrary signs and internal rules.29 The definition determines the endeavour, which is to explain language’s arrangement or internal logic. This calls for “the exclusion of everything that is outside its organism or system—in a word, of everything known as ‘external linguistics,’” e.g., the relationships of language with political history, or with social institutions such as the Church and school.30 To underscore the distinction between internal and external, Saussure offers the comparison of Vossler and his school (e.g., Leo Spitzer). Vosslerites rejected positivism and the premises of Cartesian rationalism and argued that the vital feature of speech does not consist in the grammatical forms, which are shared and stable but in stylistic concretization of these abstract forms that individualize any given utterance. See Voloshinov, Philosophy of Language, supra note 5 at 51. Bakhtin found much to admire in Vossler, but Bakhtin’s different focus on the social, historical, and dialogical dimensions of language, together with his rejection of psychologism, makes his conception of language noticeably different. For a recent, masterful reconstruction and defence of the expressive view of language see Charles Taylor, The Language Animal: The Full Shape of the Human Linguistic Capacity (Belknap Press, 2016). Taylor’s book has the great virtue of showing how the objectivist position remains alive in contemporary analytic philosophy, cognitive theory, and pragmatism. 28. According to Saussure, “In separating language [langue] from speaking [parole] we are at the same time separating: (1) what is social from what is individual; and (2) what is essential from what is accessory and more or less accidental.” See Ferdinand de Saussure, Course in General Linguistics, ed by Charles Bally & Albert Sechehaye in collaboration with Albert Riedlinger, translated by Wade Baskin (Philosophical Library, 1959) at 14. 29. Ibid. Saussure states that “[t]he linguistic sign unites, not a thing and a name, but a concept and a sound-image” (ibid at 66). The first he calls signified [signifié] and, the second, signifier [signifiant]. See ibid at 67. Their link is arbitrary because the bond between one and the other is unmotivated or without natural connection (i.e., it could be represented just as well by any other sequence). See ibid at 65-70. 30. Ibid at 20. “[I]ndeed,” Saussure writes, “the science of language is possible only if the other elements are excluded” [pas mêlés]” (ibid at 15). 440 (2022) 59 OSGOODE HALL LAW JOURNAL chess.31 To illustrate, if one begins to use ivory chess pieces instead of wooden ones, the change will have no effect within the system; in contrast, an increase in the number of chess pieces will have a profound effect on the “grammar” of the game. As a rule of thumb, then, “everything that changes the system in any way is internal.”32 And just as the game of chess is entirely comprised of the combination of the different pieces on the chessboard, language-as-system is similarly an “interdependent whole,” where each element holds each other in equilibrium in accordance with fixed rules.33 The objects of linguistic study are signs and their relations in an endless chain of substitutions34 within a pre-given system that individuals must passively assimilate.35 Saussure’s definition of language closely resembles the positivist concept of law as a system of rules. This positivist concept dominated twentieth-century jurisprudence, and as we will see in Section VIII, below, even Saussure’s metaphor of chess has been famously adopted.36 Like Saussurean linguistics, legal positivism draws a strict boundary between internal and external, law and non-law. The task of the positivist-minded legal scholar is to describe the system’s internal rules, its institutions and principles; and everything external to the system, e.g., history, politics, and morality, should be treated as not “properly legal.”37 Among the things excluded by Saussure and legal positivists alike are value judgements. In Saussurean linguistics (as in fields that share the same positivist presuppositions), value is a purely systemic or functional term, like factors in a mathematical equation or pieces in the game of chess, whereby the “value” of each piece “depends on their position on the chessboard”—and “above all else on an unchangeable convention, the set of rules that exists before a game begins and persists after each move.”38 When Saussure describes language as a system of 31. Ibid. The comparison is not fanciful, for “of all comparisons that might be imagined, the most fruitful is the one that might be drawn between the functioning of language and a game of chess” (ibid at 88). See also Section VIII, below. 32. Saussure, supra note 28 at 23. 33. Ibid at 113, 110. 34. Ibid. According to Saussure, “Language is characterized as a system based entirely on the opposition of its concrete units” (ibid at 107). Hence, the often-repeated statement about which much has been made by post-structuralism that “in language there are only differences” (ibid at 120). 35. Ibid at 14. 36. Chess is one of the favourite metaphors of HLA Hart (see Section VIII, below). 37. Most famously, see Hans Kelsen, Pure Theory of Law (1934), 2nd revised & enlarged German ed, translated by Max Knight (University of California Press, 1967). 38. Supra note 28 at 88. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 441 “pure values,”39 he means that substantive (i.e., ideological, cultural, emotional, social) values have been evacuated and replaced with their systemic or functional equivalents. It is in this sense that language can be “a form, and not a substance.”40 Lastly, Saussure distinguishes the task of describing a language system synchronically from the task of explaining its evolution diachronically. This gives way to a bifurcated field—one studying the internal or logical relations binding different terms together (synchronic or static) and the other studying their succession in time (diachronic or evolutionary). The researcher is to choose either one or the other, because the principles, perspectives, and methods of the respective approaches are mutually irreducible and antinomical.41 The distinction between synchrony and diachrony is also popular among positivist legal theory, where the static study of the system—its rules, principles, and structures—usually prevails over the evolution, often thought of as “external” to the law, e.g., having its cause in the realms of politics, economy, et cetera.42 In sum, Saussure conceives of language as a “self-contained whole,”43 an interrelated system of differences based on arbitrary signs. As a form without substance, language is neutral or indifferent to value. The focus of Saussurian linguists is the normative and self-identical features of language (grammar, syntax, semantics), excluding what is outside it. Such a linguist can study either synchronic unity or diachronic evolution, but not both. Bakhtin not only rejects each and every one of these premises but repudiates the very grounds upon which they stand, arguing that there is “no real moment in time when a synchronic system of language could be constructed.”44 For Bakhtin, “language…is never unitary,” or, “[i]t is unitary only [if ] taken in isolation from the concrete, ideological conceptualizations that fill it, and in isolation from the uninterrupted process of historical becoming that is characteristic of all language.”45 Language is not unitary in another sense too, for it contains a plurality of (stratified) “languages” within itself: the lingos and jargons of different professions; of different generations; of schools of thought; of institutions; of peoples and groups; of diverse communities; and so on. Ibid at 111. Ibid at 122, 113. Ibid at 91, 99-100. See e.g. Norberto Bobbio, A Theory of the Legal Order (Giappichelli, 1960). Supra note 28 at 9. For a criticism, see Voloshinov, Philosophy of Language, supra note 5 at 78-82. 44. Voloshinov, Philosophy of Language, supra note 5 at 66. 45. Bakhtin, “Discourse,” supra note 1 at 288. 39. 40. 41. 42. 43. 442 (2022) 59 OSGOODE HALL LAW JOURNAL The focus on abstract grammatical forms can be useful for some purposes (e.g., in language instruction), but from a Bakhtinian perspective, it “cannot serve as a basis for understanding and explaining linguistic facts as they really exist and come into being.”46 Contrary to what Saussure asserted, individuals do not encounter a ready-made system of language they must passively assimilate; rather, they enter streams of communication, where linguistic forms are being adjusted to fit each particular context.47 As a result, the primary experience of language is not an abstract system of linguistic forms but rather the social event of communication,48 in which the important thing for speakers is not the recognition of linguistic forms but their understanding of particular, concrete utterances.49 Bakhtin criticized the tendency among linguists in the objectivist vein—and the same occurs in legal positivism—to reify their object of study and then posit the characteristics of such arrested object (formal, impersonal, a priori, closed) as its essence.50 “In reifying the system of language…abstract objectivism makes language something external to the stream of verbal communication,” when in fact “language moves together with that stream and is inseparable from it.”51 The very distinction between synchrony and diachrony, as if they were two different states of the same primal element, is deceptive: This gives the impression that one could construct a static image of language independent from its concrete history, severed from its inner life and dynamics, and then explain its evolution as a succession of yet another set of static images.52 The severing of language from its lived context finds its roots in the study and decipherment of dead languages, historically, “a ready-made and handed-down body of authoritative thought,” or “someone else’s now voiceless Voloshinov, Philosophy of Language, supra note 5 at 82. Ibid at 68. Ibid at 94. Ibid at 68. On reification as the conversion of a human concept into an external thing, whereby a reified law is said to “strive,” “desire,” “make claims,” “aspire,” et cetera, see Desmond Manderson, Songs Without Music: Aesthetic Dimensions of Law and Justice (University of California Press, 2000) at 160ff. See also Davies, supra note 19 at 30 (arguing that “[l]aw does not do anything or say anything itself, and it is not even an identifiable thing—all these are shorthands for the actions of human beings enmeshed in material contexts who use an imaginary of law to relate and engage”) [emphasis in original]. 51. Voloshinov, Philosophy of Language, supra note 5 at 81. 52. See also Boris Gasparov, Beyond Pure Reason: Ferdinand de Saussure’s Philosophy of Language and Its Early Romantic Antecedents (Columbia University Press, 2013) at 117. 46. 47. 48. 49. 50. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 443 word.”53 This hypothesis finds an apt legal parallel in the reception of Roman law after the rediscovery of Justinian’s Corpus Juris Civilis in the eleventh and twelfth centuries in Bologna, which led to an enormous body of commentary and the systematization of an alien and dead (reified) law throughout the Western hemisphere—nothing like the living, actual Roman law of its time.54 Bakhtin did not try to deny the strong social currents of linguistic uniformity, which make their presence felt in norms for “correct” expression, spelling, pronunciation, and grammar.55 But he noted countervailing currents as well: Those that innovate, challenge, disrupt, sidestep, or directly confront the hegemonic control of linguistic norms. Moreover, norms of grammar are not above the fray of social contestation: Think, for example, of contemporary battles over the use of inclusive pronouns, including movements to adopt the singular “they” in English. In light of this linguistic reality, Bakhtin draws a distinction between the centripetal and centrifugal forces of language. The former “serve[s] to unify and centralize the verbal-ideological world”;56 the latter expresses the lines of flight from a unitary center. The codification of language is the theoretical expression of the centripetal or unifying forces of language, but these forces operate in the midst of “heteroglossia,” a Bakhtinian neologism that alludes to the inner pluralism of every language. As Bakhtin wrote, “a unitary language is not something given [dan] but is always in essence posited [zadan]—and at every moment…it is opposed to the realities of heteroglossia.”57 Here, too, an analogy can be drawn between the centripetal forces of language and the legal codification process of the nineteenth century, opposed by the historical school of Savigny on the basis 53. Voloshinov, Philosophy of Language, supra note 5 at 78 [emphasis added]. In Nietzschean vein, Voloshinov writes: “Despite the vast differences in cultural and historical lineaments from the ancient Hindu priests to the modern European scholar of language, the philologist has always and everywhere been a decipherer of alien, ‘secret’ scripts and words,” that is, “always and everywhere priests” (ibid at 74). 54. Goodrich, Legal Discourse, supra note 6 at 33. See also Harold J Berman, “The Origin of Western Legal Science in the European Universities” in Law and Revolution: The Formation of the Western Legal Tradition (Harvard University Press, 1983) 120. 55. Bakhtin, “Discourse,” supra note 1 at 270. For grammar as a project of political uniformization at the service of Empire, see the fascinating account of Ivan Illich about Antonio de Nebrija’s project to build the first grammar of any modern European tongue (Castilian) in 1492, offered to Queen Isabella. See Ivan Illich, “Vernacular Values” in Lawrence J Trudeau, ed, Literature Criticism From 1400 To 1800: Volume 251 (Gale, 2016) 148. 56. Bakhtin, “Discourse,” supra note 1 at 270. 57. Ibid at 270. 444 (2022) 59 OSGOODE HALL LAW JOURNAL of the living forces of law or the Volksgeist. Understood as “generative forces” rather than as abstract imperative norms, the centrifugal and centripetal also resemble the paideic and the imperial forces of law described by the non-positivist legal theory of Robert Cover, the one fostering reproduction and the other social cohesion, which operate simultaneously to guarantee the continued life of every normative universe or nomos.58 Another of Saussure’s key points is the arbitrariness and conventionality of linguistic signs, which means that language is pure form and no substance. For Bakhtin, on the other hand, language can be arbitrary to the extent that it is the result of forces (historical, cultural, sociological) that no one can master; but these leave an imprint that can be registered and traced in the signs of language, though often in a “refracted” rather than direct manner. Signs, and their embodiment in words, are “the most sensitive index of social changes,” sometimes reflecting, other times refracting, these changes.59 Bakhtin suggested that words have not only breadth or scope (the range of meanings sometimes called their “extension”) but also depth.60 Think, for instance, of the profundity (and not just the polysemy) of a word like “man”—in Sophocles’s “Ode to Man” in Antigone; in the Book of Job, during the Italian Renaissance; in Primo Levi’s account of concentration camps, If This is a Man; or in contemporary feminist and post-humanist literature.61 Words not only have histories but are ideologically charged: They come inflected with the opinions and judgments of multitudes. We inherit the words we use, but we can also reaccentuate and inflect them with our own opinions and judgments. As Bakhtin wrote, “Any concrete discourse [or utterance] finds the object at which it was directed already as it were overlain with qualifications, open to dispute, charged 58. Robert M Cover, “The Supreme Court, 1982 Term—Foreword: Nomos and Narrative” (1983) 97 Harv L Rev 4. For an explanation, see Julen Etxabe, “The Legal Universe after Robert Cover” (2010) 4 Law & Humanities 115 at 125-29. 59. Voloshinov, Philosophy of Language, supra note 5 at 19. Voloshinov states that: [T]he word is the most sensitive index of social changes, and what is more, of changes still in the process of growth, still without definitive shape and not as yet accommodated into already regularized and fully defined ideological systems…The word has the capacity to register all the transitory, delicate, momentary phases of social change [emphases in original]. The word “refracted,” is a term employed both by Voloshinov and by Bakhtin. See ibid at 24; Bakhtin, “Discourse,” supra note 1 at 302, 324, 332. 60. See Bakhtin, “Text,” supra note 24 at 127. Bakhtin speaks of “[t]he microworld of the word” (ibid). 61. For an attempt to bypass it, see Donna Haraway, Staying with the Trouble: Making Kin in The Chthulucene (Duke University Press, 2016). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 445 with value, already enveloped in an obscuring mist—or, on the contrary, by the ‘light’ of alien words that have already been spoken about it.”62 As a result, no member of the linguistic community can ever find words that are neutral, nor can language be ever rendered pure. While dictionaries give the impression that words are impersonal (i.e., spoken by no one in particular), words can only “make sense” and be understood in specific contexts and when expressed by actual interlocutors. Thus, Bakhtin distinguished between a word’s linguistic signification—the abstract and impersonal descriptions found in dictionary entries—and its actual meaning—its concrete situation in the event of communication.63 While the former ensures recognition of linguistic symbols, the latter is always a unique event.64 Abstract objectivism seeks to isolate the identity factor of the word from the multiplicity of contexts in which it can appear. That is, it seeks to remove the word from the situation in order to give it a definition outside of context—to create a “dictionary meaning out of it.”65 This is often accompanied by a move that further contributes to the reification of meaning, namely, to conceive of a central or core instance of the object to which the word refers. In doing so, abstract objectivism seeks to determine a core semantic content or central meaning of the word in question, to which all other possible meanings remain peripheral or marginal. We can find echoes of this in the jurisprudential arena, for example, in Hart’s concepts of core and penumbra.66 But according to Bakhtin, the discrimination between central and peripheral, usual and occasional, rests on the fallacy “of presupposing that that aspect really does exist and is stable.”67 Such an assumption may be justified only from the standpoint of language instruction (or deciphering a dead and alien language), but cannot explain the dynamic reality of language as it appears in living speech; it does not consider the perspective of the speakers immersed in the flow of speech communication. In this Bakhtinian framework, it is impossible to convey the meaning of a particular word without constructing it as part of an example utterance—that 62. Bakhtin, “Discourse,” supra note 1 at 276. 63. Ibid at 281. For a discussion, see Voloshinov, Philosophy of Language, supra note 5 at 78-80. For terminology, see Todorov, The Dialogical Principle, supra note 23. 64. Bakhtin, “Speech Genres,” supra note 24 at 88. 65. Voloshinov, Philosophy of Language, supra note 5 at 80. 66. This is related to what, borrowing from Friedrich Waissman, Hart called the “open texture of language.” For a discussion, see Brian Bix, “H.L.A. Hart and the ‘Open Texture’ of Language” (1991) 10 Law & Phil 51. 67. Voloshinov, Philosophy of Language, supra note 5 at 102. 446 (2022) 59 OSGOODE HALL LAW JOURNAL is, without actualizing it.68 Thus, meaning does not belong to the word as such but to the word in its position between speakers.69 As he puts it, a word is “a bridge,” “a territory shared…by the speaker and [another] interlocutor.”70 This does not necessarily entail a meeting of the minds, because contexts of usage often clash.71As explained below, an addressee reacts to the speaker’s word; agrees or disagrees with it; adds something from the addressee’s own perspective; and answers it. As a result, the process of understanding is aimed neither at deciphering what the words themselves mean outside any context (textualism) nor at what the speaker had in mind (intentionalism). Rather, “to understand another person’s utterance means to orient oneself with respect to it, to find the proper place for it in the corresponding context.”72 Understanding is hence dialogical, which is not an expression of duality but one of plurality: The greater and deeper the capacity for response, the greater and deeper the understanding.73 We will return to Bakhtin’s all-important notion of the dialogical (and its cognates “dialogism” and “dialogized” discourse), which does not refer exclusively to face-to-face verbal encounters between two individual interlocutors, nor to a mere compositional structure of speech.74 For the moment, two different dimensions of the dialogical must be kept in mind, especially as they relate to the legal context. One of these dimensions involves other words already spoken about the same subject (“the alien word”); and the second dimension involves the responses that have yet to come (“the answering word”). Regarding the former dimension, every word we utter enters an agitated and tension-filled environment of alien words with which it weaves in and out of complex interrelationships: It merges with some, recoils from others, intersects with yet a third group, 68. Ibid at 100. 69. Ibid. As Voloshinov writes, “Meaning is the effect of interaction between speaker and listener produced via the material of a particular sound complex” (ibid at 102-03 [emphasis omitted]). 70. Ibid at 86. Throughout the article I have tried to avoid the use of gendered English, including slight adjudgment to some of the translations, except when this could not be done without substantial alteration. See the text accompanying infra notes 83 and 156, below. 71. Ibid at 80. 72. Ibid at 102 [emphasis added]. 73. Ibid. 74. Ibid. Bakhtin explains that dialogical interaction comprises not only face-to-face encounters but all sorts of verbal and non-verbal communication, including bodily gestures (e.g., raising a fist, raising the eyebrow) and tone (e.g., higher or lower pitch, warm or stern timbre). Communication can also occur silently between participants who are at a distance from one another. For example, a book, which is a “verbal performance in print,” is also an element of such communication (ibid at 95). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 447 et cetera. This means that our every word is “shaped in dialogic interaction with an alien word that is already in the object.”75 Regarding the latter dimension, “every word is directed toward an answer and cannot escape the profound influence of the answering word that it anticipates.”76 As Bakhtin wrote eloquently, “[t]he word in living conversation is directly, blatantly, oriented toward a future answer-word: it provokes an answer, anticipates it and structures itself in the answer’s direction.”77 Even further, “[f ]orming itself in an atmosphere of the already spoken, the word is at the same time determined by that which has not been said but which is needed and in fact anticipated by the answering word.”78 The word takes stock of this anticipated response—which may be imagined, invited, entreated, or presupposed—and orients itself towards it, guarding itself from unwanted interpretations, shielding itself from possible retorts, interpreting its silences, inviting sympathetic reception, et cetera. This means that the answering word is able to inscribe itself in our discourse before a response is actualized. Before elaborating on these ideas more closely, two key considerations on communication and dialogical exchanges are in order. On the one hand, it is important to distinguish Bakhtin’s conception of communication from a communicative model like Roman Jakobson’s, according to which a sender sends a message to a receiver that the latter would have to decode—which is also a prevalent model in much jurisprudential work on legal and statutory interpretation, where the task of the interpreter consists of replicating the content that the legislature communicates (see Section V, below).79 As Tzvetan Todorov explained, Jacobson’s way of conceiving of the act of communication “carries the risk of making us see linguistic exchange in the image of something like the work of telegraph operators: one person has a content to transmit, and encodes it with the help of a key and transmits it through the air; if contact is established, 75. 76. 77. 78. 79. Bakhtin, “Discourse,” supra note 1 at 279. Ibid at 280. Ibid. Ibid. Most current statutory interpretation has moved from trying to decipher what the legislature intended, towards a more objective determination of the public meaning of the legislative text, often called the “communicative content” of the legislative text. For a critique of such communicative models in jurisprudence, see Mark Greenberg, “Legislation as Communication? Legal Interpretation and the Study of Linguistic Interpretation” in Andrei Marmor & Scott Soames, eds, Philosophical Foundations of Language in the Law (Oxford University Press, 2011) 217. 448 (2022) 59 OSGOODE HALL LAW JOURNAL the other decodes it with the same key, thus recovering the initial content.”80 By contrast, “in living speech, messages are, strictly speaking, created for the first time in the process of transmission, and ultimately there is no code.”81 On the other hand, to say that language is based on dialogical interactions is not meant to cast aside or fend off conflict, which is a widespread but naïve view of dialogue that Bakhtin himself never entertained.82 Rather, the opposite: Dialogical interactions in the Bakhtinian sense are sites of struggle and expressions of power dynamics. In a long passage that deserves citation in full, Bakhtin described how he envisioned these interactions (the gendered pronoun is in the English translation): As a living…thing the word in language is half someone else’s. It becomes “one’s own” only when the speaker populates it with his own intentions, his own accent, when he appropriates the word, adapting it to his own semantic and expressive intention. Prior to this moment of appropriation, the word does not exist in a neutral and impersonal language (it is not, after all, out of a dictionary that the speaker gets his words!), but rather it exists in other people’s mouths, in other people’s contexts, serving other people’s intentions: it is from there that one must take the word, and make it one’s own. And not all words for just anyone submit equally easily to his appropriation, to his seizure and transformation into private property: many words stubbornly resist, others remain alien, sound foreign in the mouth of the one which appropriated them and who now speaks them; they cannot be assimilated into his context and fall out of it; it is as if they put themselves in quotation marks against the will of the speaker. Language is not a neutral medium that passes freely into the private property of the speaker’s intentions; it is populated—overpopulated—with the intentions of others. Expropriating it, forcing it to submit to one’s intentions and accents, is a difficult and complicated process.83 The first thing to notice is the fierceness with which Bakhtin described these processes of exchange, using coercive verbs such as “appropriating,” “seizing,” “submitting,” “forcing,” “assimilating,” and “expropriating.” The insistent use of the language of property brings to mind the zero-sum game by which property claims are often settled. The analogy further serves to convey that dialogical interactions are not cooperative—with speakers jointly participating in the creation of a shared community of language—but competitive, whereby they struggle to make words bend to their will. It would be a mistake to think that such exchanges play out in a neutral medium (e.g., the free exchange of ideas). 80. Todorov, The Dialogical Principle, supra note 23 at 55. Two additional elements are “context” and “contact (ibid at 54). 81. Ibid at 56 [translation slightly modified]. 82. Bakhtin, “Discourse,” supra note 1 at 368. 83. Ibid at 293-94 [emphasis added]. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 449 Instead, Bakhtin conceived of this process in terms of multi-directional and opposing forces, not so much as social Darwinism (a Hobbesian war of all against all) as a Gramscian struggle for ideological hegemony.84 The argument is also reminiscent of Rudolf von Jhering’s criticism of Savigny’s romantic conception of the volksgeist, which Jhering viewed in terms of a never-ending and constant “struggle for law.”85 In Bakhtin’s account, speakers may try to bend language to their will, but this can be thwarted, first of all by the intentions of other speakers. Bakhtin’s social universe is one of radical pluralism, populated by many speakers with conflicting interests and goals. Further, the attempt can be thwarted by the materiality of language itself, which can oppose “resistance.” Even when one manages to speak someone else’s words (for instance, by quoting them), they may still “sound foreign” in the mouth of the person now using them. Note that while quotation marks serve usually to bring someone else’s words into one’s own discourse, here it is the words that put themselves in quotation marks “against the will of the speaker.”86 Granting agentic capacity to words, Bakhtin referred to their “stubbornness,” as if they could and do erect barriers around themselves and refuse to be (ab)used by speakers. Notwithstanding this “stubbornness,” the agentic capacity of words operates only “as if ” they are independent from the intentions, dispositions, positions, and skills of particular speakers. Bakhtin wrote about attempts to make language “one’s own” in scare quotes, in recognition that words can never really be owned—not by anyone.87 This points ultimately to the inadequacy and limits of the property metaphor: Language cannot be objectified and turned into another commodity because it does not “belong” to anyone.88 As a “communal good,” language cannot be entirely appropriated, controlled, exhausted, or fully mastered—which is an important lesson to keep in mind for legal language too, as we will see. While Bakhtin opposed the idea of language as a neutral medium, the Saussurean notion of “pure form” has continued to be widely adopted by legal positivists and analytical philosophers. For them, language—like law itself—is 84. Brandist, The Bakhtin Circle, supra note 23. See also Craig Brandist, “The Official and the Popular in Gramsci and Bakhtin” (1996) 13 Theory Culture & Society 59. 85. Rudolph von Jhering, The Struggle for Law, 5th German ed, translated by John J Lalor (Callahan & Co, 1879). More recently, Foucault criticizes Habermas for having forgotten power relations in his communicative model. See Michel Foucault, “The Subject and Power” (1982) 8 Critical Inquiry 777 at 786-87 . 86. Bakhtin, “Discourse,” supra note 1 at 294. 87. Ibid. 88. Bakhtin, “Speech Genres,” supra note 24 at 85. 450 (2022) 59 OSGOODE HALL LAW JOURNAL an empty container that can be filled with almost any content whatsoever, and the task of an analyst is to describe this form without evaluating or influencing its content. The link here with the conventional, positivist conception of lawyers and lawyering is clear, and it is precisely this entrenched view of language and law that this article hopes to displace. The next part below begins this process by contrasting Bakhtin’s notion of the utterance with the positivist reliance on the sentence as a discrete unit of analysis. II. THE UTTERANCE VERSUS THE SENTENCE “[Human beings] speak in utterances and not in individual sentences.”89 Whereas Saussure gave up utterances as individual acts of parole, Bakhtin put them center stage as complex interactive events of communication—and hence as eminently social.90 For Bakhtin, language could exist only in the form of utterances,91 which means that utterances constitute the actual, concrete, and living reality of language.92 However, Bakhtin’s emphasis on the utterance was not an attempt simply to reverse Saussure’s langue/parole binomial. Bakhtin’s notion of the utterance is distinct: It is not just a particular act of speech but a link in the chain of speech communication. This separates Bakhtin not only from Saussure but also from writers in the expressivist tradition of Humboldt who emphasize subjective expression.93 Three important dimensions underscore the social aspect of utterances. First, they are socially grounded. Unlike abstract grammatical sentences, utterances are not free floating. An utterance does not exist in the air, in an impersonal void or ether, or out of space and time; it is performed in specific circumstances, at a certain historical juncture, and as an event of communication between speakers who are close or distant in space and time.94 These are all “essential 89. Ibid at 78. 90. The social dimension also separates Bakhtin from the subjectivist (or expressivist) school of Vossler. See supra note 27. 91. Bakhtin, “Speech Genres,” supra note 24 at 71. 92. Voloshinov, Philosophy of Language, supra note 5 at 93. 93. See supra note 27. 94. See M M Bakhtin & P N Medvedev, The Formal Method in Literary Scholarship: A Critical Introduction to Sociological Poetics, translated by Albert J Wehrle (Johns Hopkins University Press, 1978) at 120 [Medvedev, The Formal Method]. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 451 constitutive part[s]”95 from which no utterance can be totally unmoored: “When we cut the utterance off from the real grounds that nurture it…all we have left is an abstract linguistic shell or an equally abstract semantic scheme.”96 This means that “[t]he concrete utterance (and not the linguistic abstraction) is born, lives, and dies in the process of social interaction between the participants of the utterance.”97 “Language acquires life and historically evolves precisely there [at the point of interaction]…and not in the abstract linguistic system of language, nor in the individual psyche of speakers.”98 A second aspect of utterances that makes them social and not idiosyncratic is that they often develop into or within relatively stable categories, referred to by Bakhtin as “speech genres.”99 These genres vary in flexibility, significance, and scope, but tend to cast speech in generic forms such as greetings, farewells, congratulations, eulogies, all kind of wishes, information about health, business reports, and so on. To this list, we might add various legal genres and repertoires (e.g., contracts, conciliation agreements, wills, statutes, court opinions, et cetera). Speech genres guide utterances along known scripts and conventions, as well as intonational registers (e.g., “once upon a time,” “whereas,” “nevertheless”).100 They may also respond to social hierarchies, as in language etiquette and norms of speech tact. While speech genres are generally more malleable than grammatical forms, they do have “normative significance” regarding the capacity to constrain individual speakers. All this means that “the single utterance, with all its individuality and creativity, can in no way be regarded as a completely free combination of forms of language.”101 The third social dimension of utterances is that they are closely related to other utterances: They are filled with echoes and reverberations of preceding utterances (i.e., the “alien word”) and generate echoes and reverberations within those that follow (i.e., the “answering word”). To use Bakhtin’s suggestive 95. Valentin Voloshinov, “Discourse in Life and Discourse in Art” (1926), in Neal H Bruss, ed, Freudianism: A Critical Sketch, translated by I R Titunik (Indiana University Press, 1987) at 100 [Voloshinov, “Life and Art”]. 96. Ibid at 105. 97. Ibid. 98. Voloshinov, Philosophy of Language, supra note 5 at 95. 99. “Speech Genres,” supra note 24 at 60 [emphasis in original]. For Bakhtin, “[t]he wealth and diversity of speech genres are boundless, because the various possibilities of human activity are inexhaustible, and because each sphere of activity contains an entire repertoire of speech genres” (ibid). 100. For the improbable, crucial, legal significance of a simple adverb such as “nevertheless,” see Otto Preminger, The Anatomy of a Murder (Columbia Pictures, 1959). 101. Bakhtin, “Speech Genres,” supra note 24 at 81 [emphasis in original]. 452 (2022) 59 OSGOODE HALL LAW JOURNAL terminology, utterances are “aware of each other.”102 Each utterance refutes, affirms, supplements, and relies on others; presupposes them to be known; and somehow takes them into account. “[The speaker] is not, after all, the first speaker, the one who disturbs the eternal silence of the universe.”103 An utterance is also connected with subsequent utterances, to which it imagines reacting in particular ways—anticipating a retort, pre-empting criticisms, addressing an objection. Thus, an utterance is linked not only to the present (i.e., the interactive context, the pragmatic circumstances, and the broader setting) but to the past (of preceding utterances) and the future (in projections and anticipations). Arguably, then, “[t]he utterance proves to be a very complex and multiplanar phenomenon if considered not in isolation and with respect to its author (the speaker) only, but as a link in the chain of speech communication and with respect to the other, related utterances.”104 According to Bakhtin, the utterance as a “unit of communication” has not been sufficiently distinguished from the sentence as a “grammatical unit.”105 The sentence as a concept is abstracted and isolated from context and hence free-standing: A sentence can have no contact with an interactive event or broader social settings; nor with what Bakhtin called “prehistory,” the earlier lines of thought and argument spoken about the same subject;106 nor with their future, the way they will be received by the answering word. Furthermore, a sentence can express a relatively complete thought, but its particular sense of completeness is merely grammatical.107 In Bakhtin’s example, the sentence “[t]he sun has risen” is perfectly understandable (i.e., we recognize its linguistic signification and its possible role in discourse).108 Still, the sentence lacks the capability of determining any active responsive position: One cannot “agree or disagree with it, execute, it, evaluate it, and so on.”109 By contrast, an utterance—which can be formed by a single word or by an entire treatise110—expresses a position to 102. 103. 104. 105. 106. 107. 108. 109. 110. Ibid at 91. Ibid at 69. Ibid at 93. Ibid at 73. On the key differences between utterances and sentences, see Tzvetan Todorov, “Theory of the Utterance” in The Dialogical Principle, supra note 23, 41. See also Susan Petrilli, “Communication, Dialogue, and Otherness in Mikhail Bakhtin’s Metalinguistics” (2008) 1 Russian J Communication 266. Bakhtin, “Speech Genres,” supra note 24 at 73. Ibid at 74. Ibid at 82. Ibid at 74. Ibid. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 453 which one can respond.111 The possibility of response turns the sentence into an utterance112 and gives it a distinctive sense of completion, when a speaker pauses and “relinquishes the floor.”113 Thus, “The sun has risen, it is time to get up,” can be followed by “[y]es, it really is time,” or by “But it’s still very early. Let’s get some more sleep.”114 In addition to these social aspects that distinguish utterances from sentences, three key features of Bakhtinian utterances must be emphasized: addressivity, responsiveness, and evaluative intonation. While a grammatical sentence has no particular marker that inherently identifies who speaks it or to whom it is addressed—indeed, a sentence has “no author”115 or specific audience—every utterance by definition is spoken by someone and addressed to someone.116 Every word we utter is oriented towards an addressee—more precisely, towards someone particular that the addressee might be: a fellow member of the same social group, someone connected with the speaker by close personal ties, someone sharing the same cultural presuppositions, someone with whom the speaker disagrees, et cetera. For Bakhtin, the “orientation toward the listener is an orientation toward specific conceptual horizon, toward the specific world of the listener,” which entails that “[t]here can be no such thing as an abstract addressee.”117 That is to say, a speaker “senses” the listener (e.g., their familiarity or knowledge of the situation, their convictions and prejudices, their sympathies and antipathies), and adjusts the speech accordingly.118 Bakhtin’s key insight is that this “sensing”—and concomitant openness and attunement119—goes in both directions. This means that addressivity—a kind of “thrownness toward the other”120—takes us further than modern speech-act theory in incorporating the addressee into the utterance. As Bakhtin himself noted, “When constructing my utterance, I try actively to determine this [addressee’s] response,” but at the same time, “I try to act in accordance with the response I anticipate,” which “exerts an influence upon my utterance (I parry 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. Ibid at 72. Ibid at 76 [emphasis in original]. Ibid at 71. Ibid at 83. Ibid at 84. Ibid at 95. According to this, even the cry of a nursing infant is addressed to someone. See Voloshinov, Philosophy of Language, supra note 5 at 87. Ibid [emphasis added]. Voloshinov, Life and Art, supra note 95. On attunement, see Becker, supra note 2. Thomas Kent “Hermeneutics and Genre: Bakhtin and the Problem of Communicative Interaction” in Frank Farmer, ed, supra note 23, 33 at 42. 454 (2022) 59 OSGOODE HALL LAW JOURNAL objections that I foresee, I make all kinds of provisos, and so forth).”121 In fact, the addressee whom the speaker envisions makes the former’s presence felt, immanently, in the social event of communication, conditioning the utterance from within.122 Think, for instance, of how relations of personal proximity can affect not only what we say but how we say it;123 or how we are able to be tactful and modulate our language in deference to perceived attributes of the addressee: their authority, seniority, vulnerability, et cetera.124 This means that the utterance can be seen as a form of dance or movement, as the “reciprocal relationship between speaker and listener, addresser and addressee.”125 The second key element is the “responsiveness” of utterances, which alludes to the active rather than passive understanding of someone else’s speech. For Bakhtin, understanding is always active, whereas a mere passive attribution of meaning is, properly speaking, not understanding.126 When perceiving an utterance, a hearer must take an active, responsive attitude towards it—to either agree or disagree with it (completely or partially), augment it, apply it, prepare for its execution, and so on.127 Listeners assimilate words that are to be understood into their own conceptual system and enrich it with new elements, additions, and supplements.128 This response need not be verbally expressed or articulated: It may 121. “Speech Genres,” supra note 24 at 95. 122. In an incomplete text published posthumously, Bakhtin alludes briefly to the figure of the “superaddresee,”—a “third” whose absolutely just and responsive understanding is presumed. See Bakhtin, “Text,” supra note 24 at 126. There is no need to presuppose a metaphysical entity here, which is why Michael Holquist’s interpretation of the superaddressee as “the a priori of all speech” and the closest to “something like a God concept in Bakhtin” cannot be accepted. See Bakhtin, Late Essays, supra note 23 at xviii. Bakhtin is clear that “[t]he aforementioned third party is not any mystical or metaphysical being (although, given a certain understanding of the world, he can be expressed as such)” (Bakhtin, “Text,” supra note 24 at 126-27). Rather, “he is a constitutive aspect of the whole utterance, who, under deeper analysis, can be revealed in it” (ibid). In different historical times, this superadressee has acquired different ideological forms: God, absolute truth, the court of dispassionate human conscience, the people, and the court of history or science. This is not an a priori condition existing outside discourse. For obvious reasons, too, the addressee has nothing to do with the “reading public” located outside the work. See Voloshinov, “Life and Art,” supra note 95 at 114-15. 123. Bakhtin, “Speech Genres,” supra note 24 at 96-97. 124. “No word lacks tact” (Medvedev, The Formal Method, supra note 94 at 95). 125. Voloshinov, Philosophy of Language, supra note 5 at 86. I thank James Boyd White for the image. 126. Bakhtin, “Discourse,” supra note 1 at 281. 127. Bakhtin, “Speech Genres,” supra note 24 at 68. 128. Although this is not a terminology employed by Bakhtin, it may be practical here to distinguish between hearing and listening. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 455 be quietly noted or delayed; they may acquiesce or remain silent. In fact, speakers do not expect passive understanding that “only duplicates [their] own idea[s] in someone else’s mind.”129 Rather, speakers expect “response, agreement, sympathy, objection, execution.”130 It is in this sense that, for Bakhtin, “[u]nderstanding comes to fruition only in the response.”131 Given that understanding is “imbued with response and necessarily elicits it in one form or other,”132 Bakhtin concluded that “[a]ny true understanding is dialogical in nature.”133 This conclusion goes further than J.L. Austin’s speech-acts (promising, warning, declaring, appointing, dissenting), which succeed or fail in part on the basis of a set of conventional elements (which Austin called “felicity conditions”) that do not fully acknowledge the role of the listener.134 As philosopher Maurizio Lazzarato has explained, the response that the Bakhtinian utterance awaits is active, whereas in Austin, the other is “neither autonomous nor free.”135 Further, while the Austinian performative entails “convention, institution, and something to reproduce,” for Bakhtin, comprehension is always “taking up a position, a judgment, a response—an action inside dialogical relations.”136 By and large, philosophers of language have continued to adopt a passive model of understanding, focusing on passive signification rather than responsive meaning.137 The same is largely true of theories of statutory interpretation, where the meaning of a legislative text is usually determined by the semantic or asserted content of the legislative text, without consideration of the responsiveness of 129. 130. 131. 132. 133. 134. 135. 136. 137. Bakhtin, “Speech Genres,” supra note 24 at 69. Ibid. Bakhtin, “Discourse,” supra note 1 at 282. Bakhtin, “Speech Genres,” supra note 24 at 68. Voloshinov, Philosophy of Language, supra note 5 at 102. See also Bakhtin, “Text,” supra note 24 at 125. See Marianne Constable, Our Word is Our Bond: How Legal Speech Acts (Stanford University Press, 2014) at 33. Stanley Cavell’s “passionate utterances” are neither constative nor performative but focus on the “you.” A passionate utterance such as “I insult you” or “I seduce you”—or even “I persuade you”—does not do what it says in the same way as the performative “I promise…” does and requires a hearer who is attuned to it in a particular way. See Stanley Cavell, “Performative and Passionate Utterances” in Philosophy the Day After Tomorrow (Belknap Press, 2005). For the importance of the auditory sense in law, see Parker, supra note 19. Maurizio Lazzarato, “Mikhail Bakhtin’s Theory of the Utterance,” translated by Arianna Bove (2009), online: <www.generation-online.org/p/fp_lazzarato6.htm>. Ibid. Bakhtin, “Discourse,” supra note 1 at 281. 456 (2022) 59 OSGOODE HALL LAW JOURNAL the audience.138 One of the reasons for this is that jurisprudes in the analytical tradition concentrate on sentences rather than on utterances. Even when they imagine two partners in speech-communication (“the speaker” and “the hearer”), the speaker is shown as actively sending a message, while the receiver is represented as a passive receptor, or at most as a decoder.139 We will return to this issue below in the context of law and legal language, but for now, it is sufficient to point out that this one-way diagram leaves out one of the most essential aspects of speech communication, namely the (re)active role of the listener.140 But perhaps the most crucial dimension of utterances (and the key difference from the kind of sentences analyzed by jurisprudes) is their evaluative aspect. By “evaluation,” Bakhtin did not mean concrete opinions or judgments (i.e., “what do I think about this or that”) but rather a deeper, ingrained attitude presupposed by the very fact (but also the manner) of speaking about anything. Indeed, “the mere fact that I have started speaking about an object, that I have turned my attention upon it, singled it out…means that I have already assumed an emotional-volitional position, a valuational attitude, in relation to it.”141 As Bakhtin explained, this is neither a universal valuation of an object nor a passive psychic reaction but a certain “ought-to-be” attitude constituted by my participation in a lived experience.142 In this view, everything that enters 138. See infra notes 167-168 and accompanying text. See also infra note 282. For a theory of “plain” or “literal” meaning based on a “universal” or “baseline” context that remains largely invariant across competent speakers of a given language-system, see Frederick Schauer, Playing by the Rules: A Philosophical Examination of Rule-based Decision-making in Law and in Life (Oxford University Press, 1991). 139. Bakhtin, “Speech Genres,” supra note 24 at 68-70. For an attempt to develop a more complex differentiation, see Erving Goffman, Forms of Talk (University of Pennsylvania Press, 1981) 124 (distinguishing between “animator,” “author,” and “principal”). 140. Bakhtin, “Speech Genres,” supra note 24 at 70. 141. MM Bakhtin, “Author and Hero in Aesthetic Activity” in Michael Holquist & Vadim Liapunov, eds, Art and Answerability (University of Texas Press, 1990) at 225 [emphasis in original] [Bakhtin, Art]. 142. See MM Bakhtin, Toward the Philosophy of The Act (University of Texas Press,1993) at 36 [Bakhtin, Act]. In this text dating to 1919-1921 and published posthumously, the emotional-evaluative relation with the objects of experience acquires an ontological-existential dimension that transforms mere possibility into the actuality of being-as-event (ibid at 32-45). In this view, “[e]verything that is actually experienced…has an emotional-volitional tone and enters into an effective relationship to me within the unity of the ongoing event encompassing us” (ibid at 33). By contrast, an indifferent attitude cannot “generate sufficient power to slow down and linger intently over an object, to hold and sculpt every detail and particular in it, however minute” (ibid at 64 [emphasis in original]). Answerability underlies what Bakhtin calls the “non-alibi in Being,” that is, that I cannot excuse or relieve myself from the responsibility of the act (ibid). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 457 our field of experience presupposes a certain evaluative attitude—an attitude of “non-indifference”—towards it. “[E]valuation organizes how we see and conceptualize the event being communicated, for we only see and conceptualize what interests or affects us in one way or another.”143 As a result, “evaluation… mediates between language as an abstract system of possibilities and the concrete reality of language.”144 In other words, “[o]nly through evaluation do the possibilities of language become real.”145 For Bakhtin, utterances are essentially evaluative: “that’s a lie,” “that’s the truth,” “that’s a daring thing to say,” “you can’t say that” merge with the particular event, forming with it an “indissoluble unity.”146 Hence, “[i]t is impossible to understand the concrete utterance without accustoming oneself to its values.”147 While evaluation thus defines all aspects of the utterance and totally permeates it, evaluation most commonly finds outlet in expressive intonation, which affects sound and meaning at once.148 Intonation can be manifested in subtle inflections in the voice, in bodily gestures, and in a myriad of other ways of emphasizing and qualifying the relevant experience, whether it is joy, sorrow, doubt, uncertainty, ambivalence, contempt, and so on.149 Although standard types of expressive intonation (phrases such as “Oh No!,” “Hurray!” or gestures such as raising a fist) can be identified, there is an extremely subtle and complex set of possibilities for intoning an experience.150 Intonation issues from the speaker’s evaluative attitude towards the subject-matter of speech and the communicative event (e.g., the concrete circumstances of speech, the explicit or implicit participants, the audience’s assumed or perceived reactions, 143. Medvedev, The Formal Method, supra note 94 at 127. 144. Ibid at 125. As Bakhtin writes: Insofar as we abstractly separate the content of a lived-experience from its actual experiencing, the content presents itself to us as something absolutely indifferent to value qua actual and affirmed value…Yet in order to become really actualized and thus made into a participant in the historical being of actual cognition, [it]…must enter into an essential interconnection with an actual valuation (Bakhtin, Act, supra note 142 at 33). 145. Medvedev, The Formal Method, supra note 94 at 123. 146. Voloshinov, “Life and Art,” supra note 95 at 98. 147. Medvedev, The Formal Method, supra note 94 at 121. On what this “accustoming” might mean, see Alton Becker’s notion of “attunement,” which also requires coming to terms with the silences and the unsaid in language. See supra note 2. 148. Medvedev, The Formal Method, supra note 94 at 122. 149. Arguably, “[i]ntonation always lies on the border of the verbal and the nonverbal, the said and the unsaid.” Voloshinov, “Life and Art,” supra note 95 at 102. 150. Voloshinov, Philosophy of Language, supra note 5 at 87. 458 (2022) 59 OSGOODE HALL LAW JOURNAL et cetera). A sentence like “he died” usually embodies a certain emotional register, but depending on the event, for example the death of a particularly hateful dictator, “he died” can also reflect a positive, even a rejoicing expression.151 Similarly, the expression “what joy!” can assume a sarcastic, bitter, or resigned tone of boredom. In some dramatic circumstances, an apparently descriptive word can acquire profound emotion, as in Xenophon’s account of the March of the Ten Thousand Greek soldiers, who, after three months of being lost, cut off, and surrounded by enemies in Persian territory, were able at long last to exclaim “[T]halassa! [T]halassa!” [The sea! The sea!].152 Intonation, however, is no mere subjective or emotive expression. In order to appreciate the social fabric of intonation, Bakhtin offered a little scene of two interlocutors looking up at the window and realizing to their dismay that it has started to snow in the middle of May, after which one of them says to the other “well!” In this vignette, both interlocutors are sick and tired of the protracted winter and anxious to welcome the spring, which the sudden May snow now frustrates.153 This scene is not privative to the speakers: Just about anyone who has spent a long winter in the Northern hemisphere can identify (and sympathize with) the tone of disappointment, resignation, and dry humour that is laconically contained in the expression “well!”—a word that might otherwise appear nonsensical in response to May snow. Intonation is thus not a subjective mental state; rather, it entails and brings about a social world of relations (e.g., friends, allies, opponents, neighbours).154 In fact, intonation is “especially sensitive to all the vibrations in the social atmosphere surrounding the speaker.”155 For instance, when a speaker tells a joke in an atmosphere of sympathy and has the audience’s choral support, intonation becomes rich and expansive. By contrast, when such choral support is lacking, the voice often falters and its intonational richness is reduced, “as happens for instance, when a person laughing suddenly realizes that he is laughing alone—his laughter either ceases or degenerates, becomes forced, loses its assurance or clarity and its ability to generate joking and amusing talk.”156 Therefore, when a person 151. Bakhtin, “Speech Genres,” supra note 24 at 85. 152. Ibid at 86. The story is told in Xenophon’s Anabisis (ca 370 BC) and the sea they saw was the black sea, which symbolized the possibility of returning home. See Xenophon in Seven Volumes, translated by Carleton Lewis Brownson (Harvard University Press, 1924), s I(IV). 153. The scene is described in Voloshinov, “Life and Art,” supra note 95 at 99-105. 154. Ibid at 104. 155. Ibid at 102. 156. Ibid at 103 (gendered language in the English translation). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 459 anticipates the disagreement of the interlocutors or, at any rate, is uncertain or doubtful of their agreement, the speaker intones the words differently. A first corollary from the preceding discussion is that intonation merges with the event and forms with it an “indissoluble unity.”157 In other words, intonation (like evaluation in general) is a constitutive aspect of the utterance that cannot be divorced from its meaning.158 Depending on how we intone it, an expression of hunger may alternatively express rightful indignation or a humble request; it may convey dissatisfaction with one’s cruel destiny or with society; or it may take on political coloring, as a call for political protest and agitation.159 Therefore, the usual distinction between “denotative” (referential) and “connotative” (evaluative) content, as if the former would be the primary meaning of the statement and the latter added on top of it, stems from a failure to realize the constitutive role of evaluation in living speech.160 A second corollary is this: Given that there is no such thing as a word without “evaluative accent,” “[n]o utterance can be put together without value judgment.”161 To put it simply, “[t]here can be no such thing as an absolutely neutral utterance.”162 As a case in point, for a legal scholar to write in a manner that is “dispassionate, methodical, and almost mathematical”163 is undoubtedly a matter of style, and that style is inherently a matter of evaluation. Supposedly “neutral” statements are thus another way of taking a position.164 This will be explored further in Sections VI-VII, below. In sum, sentences are: (1) abstract, (2) impersonal, (3) a-contextual, (4) self-standing, (5) passive, (6) purportedly non-evaluative (atonal), and (7) univocal (or monological). In contrast, utterances are: (1) actual, (2) spoken by and to particular speakers, (3) situated, (4) inter-connected, (5) active, (6) evaluative (and intonated), and (7) dialogical. Ibid at 98. Bakhtin, “Speech Genres,” supra note 24 at 90. Voloshinov, Philosophy of Language, supra note 5 at 87. Ibid at 105. Ibid at 103, 105. Bakhtin, “Speech Genres,” supra note 24 at 84. For the description of the style of Grotius, see Oona Hathaway & Scott Shapiro, The Internationalists: How a Radical Plan to Outlaw War Remade the World (Simon & Schuster, 2017) at 12. 164. Bakhtin conceives the possibility of neutrality (or abstraction) only at the level of grammatical sentences. See “Speech Genres,” supra note 24 at 82. But, of course, we do not speak in sentences, but in utterances. 157. 158. 159. 160. 161. 162. 163. 460 (2022) 59 OSGOODE HALL LAW JOURNAL *** The argument to be developed in the following Sections is that legal positivism and analytical jurisprudence still think in terms of sentences and not of utterances. That is, the conception of language underpinning their legal theories, as well as their characteristic ways of pursuing them, relies on the features of sentences rather than of utterances as described by Bakhtin. Legal positivism focuses on legal propositions, typically enacted by rules that are (supposedly) complete on their own and isolated from the stream of speech communication that connects utterances to the past, present, and future. Unmoored from concrete interactive events, positivists of the analytical persuasion pay little attention to the speakers and addressees of legal speech, or to the dynamic relationships established between the two. Instead, they typically favour the use of formal symbols or the conventional, dictionary, and “ordinary” meaning of words, to the exclusion of historically and culturally charged ones. In doing so, they undervalue the role of legal audiences who are not mere hearers upon whom an ideal “law” is enacted but rather dialogical participants who are active and responsive and therefore fundamental to the creation of legal meaning. This general criticism holds true despite the fact that legal positivists, under the influence of contemporary philosophy of language, have made some efforts to incorporate pragmatic and contextual considerations into their analyses.165 Yet in these undertakings, context is often understood as something added to, and in fact logically independent from, linguistic content, rather than indissolubly united and integral to meaning, as Bakhtin defined it.166 On the former view, context forms a sort of surrounding perimeter of the proposition, without penetrating its linguistic core.167 This is evident in positivist analyses of the language of legislative texts. For example, Andrei Marmor has argued that, concerning lawmaking activity, “[l]egislatures manage to say most of what they want to say with very minimal contextual presuppositions and very limited 165. Schauer, supra note 138; Marmor, Language of Law, supra note 17. See also Marmor & Soames, supra note 79. More broadly, see also Scott Soames, Philosophical Essays, Vol. 1. Natural Language: What it Means and How to Use It (Princeton University Press, 2008) [Soames, Philosophical Essays]. 166. For Marmor’s discussion on “pragmatic enrichment,” see Language of Law, supra note 17. See also S Soames, “Interpreting Legal Texts: What is, and What is not, Special about the Law” in Philosophical Essays, supra note 165, 403. For the notion of pragmatic implicatures, see Paul Grice, “Logic and Conversation” in Studies in the Way of Words (Harvard University Press, 1991) 22. 167. See e.g. Schauer, supra note 138. See Marmor, supra note 17. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 461 pragmatic enrichment,” which means that “there is no gap between the content by the legislature and the legal content of the act.”168 Arguably, legal positivism remains trapped in a “container theory of meaning” in which the central task is “to specify the truth-conditions of its sentences”169— in other words, to identify the conditions in virtue of which the content of legal propositions is true or false in a given context. Thus, legal positivists devise typologies of norms (e.g., permissions, prohibitions, power-conferring norms, principles and so on) and try to define the abstract qualities or “properties” of each.170 Like bricklayers building a wall one brick at a time, legal positivists seek to (re)construct the legal system one proposition at a time. Like taxidermists dissecting legal language, positivists pay little attention to live utterances and, in particular, to the addressees, their responsiveness, and the necessarily evaluative character of every legal utterance. Instead, the addressee is often imagined as the passive receptacle of the lawmaker’s communicative intention, tasked merely with duplicating what the law, statute, or the text in question demands, rather than contributing crucially to its actualization.171 Naturally, then, rules are prototypically exemplified as commands and exhortations,172 which demand little more than obedience.173 When the addressees of law are considered, they are conceived as abstract or notional constructs such as the “reasonable hearer,” “or the average speaker of English,”174 rather than addressees speaking from concrete positions. Similarly, when positivists make forays into the performative aspects of language they often stress the conventional or “rule-governed” aspect of language, 168. Marmor, supra note 17 at 27, 12. See also Hrafn Asgeirsson, “Authority, Communication and Legal Content” in The Nature and Value of Vagueness in Law (Hart, 2020) 6. 169. See Soames, Philosophical Essays, supra note 165 at 2. This theory can also be called “representational,” for “to know the meaning of a sentence is to know the way in which uses of it represent the world to be” (ibid). 170. See Atienza & Manero, supra note 8, with accompanying bibliography. 171. For a pragmatist critique of legislative intent from the perspective of the audience, see Barbara Baum Levenboook, “Soames, Legislative Intent, and the Meaning of a Statute” in Graham Hubbs & Douglas Lind, eds, Pragmatism, Law and Language (2013) 40. 172. This is not to say that positivists do not consider different kinds of rules. For a typology, see Atienza & Manero, supra note 8. 173. Hart’s legal theory—and in particular his most crucial rule of recognition—is in fact addressed to legal officials whose acceptance alone suffices. A thinner understanding of ordinary people’s engagement with the law is hardly imaginable. HLA Hart, The Concept of Law, 3rd ed (Oxford University Press, 2012) at 116 [Hart, The Concept of Law]. 174. For some of these categories, see Brian G Slocum, “Ordinary Meaning and Empiricism” (2019) 40 Stat L Rev 13 at 21-24. 462 (2022) 59 OSGOODE HALL LAW JOURNAL that is, its system-like qualities.175 In this way, legal positivism’s attention to the performative aspect of speech is only skin-deep.176 This leads to the conclusion, finally, that legal positivism fails to account for the evaluative character of legal language and for its own “intonation” as a legal theory that can somehow be free from value judgment. But before we can advance all of these arguments in detail, there is one last category of language that must be explained, which is the famous Bakhtinian concept of the dialogical. III. DIALOGICAL RELATIONS AND THE INTERNAL DIALOGISM OF DISCOURSE “[T]he real unit of language…is not the individual, isolated monologic utterance, but the interaction of at least two utterances—in a word, dialogue.”177 “Agreement is very rich in varieties and shadings.”178 No account of Bakhtin can be complete without an understanding of “the dialogical,” which might rightfully be Bakhtin’s key term of art—so much so that it figures in the title of one his most popular collection of essays.179 But precisely because it is a term of art which does not tally with its conventional use, and because Bakhtin himself employed it liberally without being constrained by rigid categories, it may be helpful to clarify it. Without prejudice to the myriad ways 175. See e.g. Marmor, Language of Law, supra note 17 at 23. See in particular John R Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge University Press, 1969). Searle emphasized that speech acts depend upon the realization of a series of underlying rules, characteristically performed by uttering certain expressions in accordance with them (ibid at 37). Therefore, while potentially divergent, speech-act theory remains largely compatible with the idea of language as a system of signs. In fact, Searle described his contribution explicitly in Saussurean terms, as a study not of “parole” but of “langue” (ibid at 17). 176. See Hart, The Concept of Law, supra note 173. Hart claimed to be following JL Austin’s “sharpened awareness of words to sharpen our perceptions of phenomena” (ibid at 14). He also considered Austin’s work on performatives to be of permanent value for analytic jurisprudence. See Matthew H Kramer, “Hart and the Metaphysics and Semantics of Legal Normativity” (2018) 31 Ratio Juris 396 at 406. Kramer has criticized Hart’s overemphasis on semantics in his account of legal interpretation but maintains that Hart was well attuned to the pragmatics of legal discourse in his overall theorizing about the law (ibid at 403). Kramer’s account does not satisfactorily explain why Hart’s underlying conception of language could be so different in both instances. 177. Voloshinov, Philosophy of Language, supra note 5. 178. Bakhtin, “Text,” supra note 24. 179. See MM Bakhtin, The Dialogical Imagination (University of Texas Press, 1981). Michael Holquist has also published his own book on Bakhtin entitled Dialogism: Bakhtin and His World, 2nd ed (Routledge, 2002). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 463 in which dialogue informs every aspect of everyday life, the term “dialogical” can be understood in at least three different senses: social, epistemological, and discursive.180 In the first, social sense, dialogue is part of the uninterrupted and continuous flow of communication, whereby an individual utterance, no matter how complete in and of itself, is only a moment or point in a web of social interaction.181 Therefore, “[t]he living utterance, having taken meaning and shape at a particular historical moment…cannot fail to become an active participant in social dialogue. After all, the utterance arises out of this dialogue as a continuation of it and as a rejoinder to it.”182 When Bakhtin argued that meaning and understanding proper are “dialogical,” he used the term in this sense to convey that the meaning-making process occurs in the space of social encounter. The opposite of the dialogical in its social dimension would be the abstract and decontextualized sentence, severed from the stream of communication.183 In a second, epistemological sense, the dialogical stands for the profusion of languages (including genres and styles of language) that flourish in the midst of centripetal or unifying forces. The dialogical embodies the pluralist impulse to diversify and resist the tendency towards uniformization. In this broad matrix, Bakhtin analyzed entire genres of discourse that best represent such dialogical impulse (including the polyphonic novel, Menippean satire, the carnival) and traced their historical manifestations in the sociolinguistic and ideological world.184 At the other end, centripetal forces are embodied by genres of discourse such as Aristotelian logos, the Augustinian true word, the orthodoxy of the medieval church, Cartesian rationalism, Leibniz’s idea of universal grammar, or Saussurian linguistics, all of which serve to centralize and unify language.185 In the context of law, we may think of the nineteenth-century movement of legal codification, as well as more recent projects of uniformization and legal harmonization (European law, Uniform Commercial Codes, attempts of harmonization of private law by UNIDROIT, global administrative law, 180. For a slightly different categorization distinguishing between ontological, epistemological, and metalinguistic understandings of dialogue, see Frank Farmer, “Introduction” in Frank Farmer, supra note 23, xi. 181. Voloshinov, Philosophy of Language, supra note 5 at 72, 95. 182. Bakhtin, “Discourse,” supra note 1 at 276-77. 183. Voloshinov, Philosophy of Language, supra note 5 at 72-74. 184. See especially Bakhtin, “Discourse,” supra note 1 at 273ff. See also Bakhtin, Dostoevsky, supra note 23; Bakhtin, Rabelais, supra note 23. 185. Bakhtin, “Discourse,” supra note 1 at 271. 464 (2022) 59 OSGOODE HALL LAW JOURNAL et cetera).186 Opposite the dialogical in this sense stand, not abstraction and de-contextualization, but purportedly neutral languages, to which all the rest must be adjusted or translated.187 The aim of the latter is invariably to overcome the curse of Babel to construct a lingua franca (or metalanguage).188 It is in this epistemological sense that Bakhtin referred to the “monologism” of official discourses that issue dogmatic and authoritarian declarations and ready-made truths.189 Bakhtin alludes further to monological reasoning in the human sciences that conceive of their own form of rationality as unique and reject other forms of discourse, evidence, and argument.190 In addition to the social and epistemological, there is a third sense of the dialogical that can be called discursive and is perhaps Bakhtin’s most singular use of the term. This sense of the dialogical, and its cognates “dialogism” and “internally dialogized” discourse, entails a realization that the languages we speak are never exclusively our own but instead populated by alien intentions, accents, and voices. As George Kamberelis and Karla Scott have explained, this sense of the dialogic “means not only that people use language to engage in dialogue but, more importantly, that any stretch of discourse (e.g., phrase, utterance, text) often involves multiple intersecting, transversing, and disrupting languages…drawn from alternative social and cultural home environments.”191 Therefore, utterances may contain voices that are infinitely distant, unnamed, almost undetectable (like voices of different language styles or lexical shadings), and others that resound 186. For some of these attempts, see J Cartwright, “Harmonisation Projects: Lessons from the European Experience?” (2018) 2 Latin American L Studies 163. 187. For a view criticizing the implicit assumption, see J B White, “The Language of ‘Concepts’: A Case Study” in Justice as Translation: An Essay in Cultural and Legal Criticism (The University of Chicago Press, 1990) 22 at 31 [White, Justice as Translation]. 188. For a thorough philosophical overview of the ethics of translation, see François Ost, Traduire: Défense et illustration du multilinguisme (Fayard, 2009). 189. Bakhtin contrasts these with the Socratic notions of truth, where truth is meant to emerge collectively in an interactive process. See Bakhtin, Dostoevsky, supra note 23 at 110. The dialogic notion of truth (and the human thinking about truth) has been productively picked up by Hannah Arendt. See “Introduction into Politics” in Jerome Kohn, ed, The Promise of Politics (Schocken Books, 2005) 93. 190. Bakhtin, Dostoevsky, supra note 23 at 292 (“Monologism, at its extreme, denies the existence outside itself of another consciousness”). “Monologue is finalized and deaf to the other’s response, does not expect it and does not acknowledge it any decisive force” (ibid at 293). It pretends to be the last or “ultimate word” (ibid). See also White, Justice as Translation, supra note 187 at 29 (for whom this form of rationality is “structurally coercive”). 191. George Kamberelis & Karla Dannette Scott, “Other People’s Voices: The Coarticulation of Texts and Subjectivities” (1992) 4 Linguistics & Education 359 at 365. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 465 nearby and simultaneously.192 These alien voices penetrate and pervade our language and can coexist side by side with our intended utterances and words.193 As Bakhtin affirmed, Our practical everyday speech is full of other people’s words: with some of them we completely merge our own voice, forgetting whose they are; others, which we take as authoritative, we use to reinforce our own words; still others, finally, we populate with our own aspirations, alien or hostile to them.194 Bakhtin was alert to these alien presences (real or virtual) in speech, which tend to go undetected by single-plane analyses of propositional and declaratory language.195 In fact, he invented a new methodology to register these various intersecting voices, which he called metalinguistics (or translinguistics).196 Crucially, Bakhtin did not simply suggest that an utterance is always situated in a socio-historical context; rather, he demonstrated “how various verbal-ideological perspectives and voices come into contact and interanimate one another.”197 While some authors speak of intertextuality in reference to the embeddedness of various texts within the framing discourse,198 I prefer to keep the term “dialogical relations” to underline the fact that, behind any textual manifestation of embeddedness, there are speaking positions, integral worldviews, and actual 192. See Bakhtin, “Text,” supra note 24 at 124. See also Boris Gasparov, Speech, Memory and Meaning: Intertextuality in Everyday Language (De Gruyter Mouton, 2010) at 12. Gasparov writes, “What dialogism means is that every act of speech, of any genre and mode, bears an imprint of the ‘other’—whether the ‘other’ is directly present or implied, known to the speaker directly or constructed” (ibid). 193. Bakhtin, Dostoevsky, supra note 23 at 184. 194. Ibid at 195. 195. Bakhtin, “Discourse,” supra note 1 at 279. 196. This is the term employed in Todorov, supra note 23. See also Roland Barthes, Elements of Semiology (Hill and Wang, 1968). 197. Kamberelis & Scott, supra note 191 at 367 [emphasis in original]. 198. The term “intertextuality” was coined by Julia Kristeva in her seminal article. See “Word, Dialogue and Novel” in Leon S Roudiez, ed, Desire in Language: A Semiotic Approach to Literature and Art (Columbia University Press,1980) 64. For an influential development, see Norman Fairclough, “Intertextuality in Critical Discourse Analysis” (1992) 4 Linguistics & Education 269. Bakhtin’s preferred term is “dialogical relations.” See e.g. Bakhtin, “Text,” supra note 24 at 121. 466 (2022) 59 OSGOODE HALL LAW JOURNAL voices.199 In Bakhtin’s sense, dialogic relations “are profoundly unique”200 and can be reduced neither to the purely logical (even dialectical) nor to the purely linguistic (grammatical, syntactic, or lexical-semantic).201 These relations are dia-logical; that is, they occur between various utterances or speaking positions.202 Bakhtin provided the following example: “Life is good”; “Life is not good.”203 Between these two judgments there exists a specific logical relationship: One is the negation of the other. But no dialogical relationship exists unless both judgments are embodied as two different utterances by two different speaking positions (e.g., commentary, retort, rejoinder, qualification, et cetera).204 Moreover, dialogical relations can arise between utterances that are identical in all other respects. For example, “Beautiful weather!” and the response “Beautiful weather!” can show proper dialogical relations that are not a simple echo.205 As Bakhtin suggestively put it, “agreement [can be]…very rich in variety and shadings.”206 According to Bakhtin, “a dialogic approach is possible toward any signifying part of an utterance, even toward an individual word, if that word is perceived not as the impersonal word of language but as sign of someone else’s semantic position…or…someone else’s voices.”207 Dialogism can therefore permeate every utterance, including one’s own utterance, e.g., if we speak with inner reservation or maintain a certain distance from it. Dialogical relations can also exist between two utterances separated in time and space: Any survey of the history of any scientific or philosophical question produces dialogic comparisons of the utterances of scientists who did not know anything of one another.208 199. In the glossary of The Dialogic Imagination, voice is defined as “the speaking personality, the speaking consciousness. A voice has will or desire behind it, its own timbre and overtones.” See Bakhtin, “Discourse,” supra note 1 at 434. However, terms like “personality,” “consciousness,” “will,” and “desire” would require further explanation. For Bakhtin, the voice is embodied and materially palpable in the language (or discourse) and not somewhere outside it. See Renfrew, supra note 23 at 82. 200. “Text,” supra note 24 at 124. These types of relations “constitute a special type of semantic relations, whose members can only be complete utterances…behind which stand (and in which [they] are expressed) real or potentially real speech subjects, authors of the given utterances” (ibid). 201. Ibid at 117. 202. Ibid. 203. For the example, see Bakhtin, Dostoevsky, supra note 23 at 183. 204. Ibid. 205. Bakhtin, “Text,” supra note 24 at 125. 206. Ibid. 207. Bakhtin, Dostoevsky, supra note 23 at 184. 208. Bakhtin, “Text,” supra note 24 at 124. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 467 Dialogical relationships can range from oppressive and claustrophobic (e.g., the reverential and authoritative voice that wants to be reiterated and learned by heart) to open and enriching (e.g., as in the ideal Socratic dialogue).209 In a discursive sense, the dialogical is opposed neither to abstraction and decontextualization (as in the social meaning) nor to uniformity and neutrality (as in the epistemological) but rather to the idea of hermetic and dogmatic, self-sufficient discourses, in which the presence of this “other” is excluded or cordoned off. Even though Bakhtin viewed some discourses as more dialogical than others that remain under the control of a “monological” author, he suggested that no discourse is able to close itself off entirely. Thus, Bakhtin could say without contradiction that dialogical relations are always present, even among profoundly monological works.210 One important manifestation of dialogism is “double-voiced discourse,” where a single utterance is actually inhabited by two voices, two manners of speech, two semantic and axiological belief systems.211 Here is one example drawn from Dostoevsky: Once in winter, on a cold and frosty evening…three extremely distinguished gentlemen were sitting in a comfortable, even sumptuously appointed, room inside a handsome two-story house on Petersburg Island and were occupied in weighty and superlative talk on an extremely remarkable topic. All three gentlemen were officials of the rank of general. They were seated around a small table, each in a handsome upholstered chair, and during pauses in the conversation they comfortably sipped champagne.212 At first sight, the description may seem unexceptional. In fact, the use of trite, exaggerated, and repetitive epithets (handsome, comfortable, distinguished, superlative, et cetera) could be criticized if we took the description seriously as emanating from the author’s own worldview. Upon closer inspection, however, it becomes apparent that the choice of vocabulary is not the author’s but epitomizes the worldview of the “three extremely distinguished gentlemen” themselves. Despite the apparently poor descriptive value of these epithets, each constitutes an arena in which two intonations, two standpoints, and two worldviews collide.213 If this second register is not perceived, the dynamic tension animating the passage is lost. 209. Obviously, it is also possible for Socratic dialogue to work in a more constraining and “coercive” way. 210. “Text,” supra note 24 at 125. 211. See the discussion in Bakhtin, Dostoevsky, supra note 23 at 185-203. 212. Voloshinov, Philosophy of Language, supra note 5 at 135 [emphasis omitted]. For the full example, see Fiodor Dostoevsky, “Skverny anekdot [A Nasty Story],” Vremya (1862) 11. 213. Voloshinov, Philosophy of Language, supra note 5 at 134. 468 (2022) 59 OSGOODE HALL LAW JOURNAL From the vantage of general linguistics—which is arguably the perspective of analytical jurisprudence and the doctrinal analysis of law—it is impossible to detect the essential differences between single-voiced and double-voiced discourse. This has important implications at the level of legal doctrine, as I want now to illustrate with a real example. In Hirst v. United Kingdom (No. 2) (“Hirst”), the Grand Chamber of the European Court of Human Rights (ECtHR) considered whether a blanket ban on prisoner voting, applied automatically and without regard for the individual circumstances of the prisoner or the crime, was in breach of the right to vote as enshrined in the European Convention of Human Rights.214 The case was tricky because, in electoral matters, states are granted a wide margin of appreciation, and generally speaking, courts are not in the habit of second-guessing Parliaments in this regard. Moreover, the case involved a general law, and the ECtHR is not entitled in principle to make an abstract control of laws (as a Constitutional Court might do). Once a prima facie case of interference in a Convention right has been established, the ECtHR must be satisfied that the governmental action (in this case, the UK legislation) pursues a “legitimate aim.” To that effect, the Government contended that the ban on prisoner voting enhanced civic responsibility by depriving those who have breached the basic rules of society of the right to have a say in the way the country is run.215 The applicant, on the other hand, contended that the measure took away civic responsibility, further alienating prisoners from society.216 Rather than engaging with the issues head-on, here is how the ECtHR responded: Although rejecting the notion that imprisonment after conviction involves the forfeiture of rights beyond the right to liberty, and especially the assertion that voting is a privilege, not a right (see paragraph 59 above), the Court accepts that [the UK law] may be regarded as pursuing the aims identified by the Government. It observes that, in its judgment, the Chamber expressed reservations as to the validity of the asserted aims, citing the majority opinion of the Canadian Supreme Court in Sauvé (no 2)…. However, whatever doubt there may be as to the efficacy 214. See 2005 ECHR 681 at para 82 [Hirst]. Article 3 of Protocol 1 states: “The High Contracting Parties undertake to hold free elections at reasonable intervals by secret ballot, under conditions which will ensure that the free expression of the opinion of the public in the choice of the legislature” (ibid). The Court concluded that the blanket ban exceeded any acceptable margin of appreciation and was disproportionate. 215. Ibid at para 50. 216. Ibid at para 44. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 469 of achieving these aims through a bar on voting the Court finds no reason in the circumstances of this application to exclude those aims as untenable or incompatible per se with the right guaranteed under Article 3 of Protocol No 1 [of the European Convention of Human Rights].217 A straightforward doctrinal reading of the passage is that the ECtHR accepted the legitimacy of the Government’s aims.218 But to what extent does this conclusion capture what the ECtHR is saying? More precisely, what does judicial “acceptance” mean in this particular context? In order to unpack this passage dialogically, we must begin by noting that before the ECtHR could be said to “accept” anything, it rejected two other notions that had been explicitly defended by the Government: that incarceration entails the loss of rights beyond the right to liberty, and that voting is a privilege. It did so, moreover, by referencing another passage of the judgment, where the Court sided with the applicant (“[a]s pointed out by the applicant, the right to vote is not a privilege”) and asserted that “in the twenty-first century, the presumption in a democratic state must be in favor of inclusion.”219 Furthermore, the ECtHR did not say that the aims asserted by the government were legitimate but rather that they “may be regarded” as such. This language opens up a hypothetical scenario, in a charitable reading, that the court is willing (conditionally) to entertain. Yet this hypothetical scenario is immediately undercut by the mention of the fact that the first instance Chamber judgment “expressed reservations” about the asserted aims, reservations which are augmented by the fact that the first instance judgment found “much force” in the arguments of the Supreme Court of Canada (SCC) in Sauvé v. Canada (Chief Electoral Officer) (“Sauvé”).220 Then, even when it appeared that the Grand Chamber was ready to leave doubts behind and “accept” that the ban may be regarded as pursuing the aims identified by the government, it did so by restating the doubts that remain about both their validity (as expressed in the first instance judgment) and their efficacy (“of achieving these aims through a bar on voting”).221 Ultimately, the ECtHR declared that it found “no reason…to exclude these aims as untenable or incompatible per se”222—a particularly convoluted, double negative construction that relies not 217. Ibid at para 75 [citations omitted]. 218. Colin Murray, “Playing for Time: Prisoner Disenfranchisement under the ECHR after Hirst v United Kingdom” (2011) 22 KLJ 309 at 312. Cf Plaxton & Hardy, “Prisoner Disenfranchisement: Four Judicial Approaches” (2010) 28 BJIL 101 at 119. 219. Hirst, supra note 214 at para 59. 220. 2002 SCC 68 [Sauvé]. 221. Hirst, supra note 214 at para 75. 222. Ibid. 470 (2022) 59 OSGOODE HALL LAW JOURNAL on the strength of the government’s position but on the inability of the ECtHR to reject it. It was not that the government’s reasons were actually compelling but that they were not “untenable or incompatible per se”—a strictly logical possibility that reveals the distance between the generic form of the argument (hypothetical, conditional, tentatively entertained) and its current evaluative orientation (full of hesitations, reservations, provisos). As if wanting to erect a wall of separation between their statement and their conviction, the ECtHR hid behind hypothetical constructions, double negatives, cross-references, and self-effacing language. By referring to someone else’s reservations and to general “doubts there may be,” the Court’s own doubts were recast in the conditional language of “acceptance.” The written judgment displays the characteristics of typical Bakhtinian double-voiced discourse. If we ask “who is doubting here?,” it is not hard to perceive the speaker behind the mask. In Hirst, the ECtHR appeared unpersuaded about the reasons behind the UK ban on prisoner voting, but it stopped short of the bold denunciation expressed by the SCC a few years earlier: “Quite simply the [Canadian] Government has failed to identify particular problems that require denying the right to vote.”223 The ECtHR, in contrast, may have been reluctant to clash head-on with the Parliament of a member state because doing so would have put it on contested philosophical, empirical, and institutional grounds. And yet it alluded to opposing value judgements in a refracted way, by referring to the arguments of others (the first instance Chamber’s reservations, the persuasive arguments of the Canadian majority in Sauvé, the doubts about the efficacy of the measures, the acknowledgements of the strength in the applicant’s position, et cetera). Without the intermediation of these other voices, the ECtHR’s full argumentation cannot be properly understood. Attending to the dialogical construction of the ECtHR’s utterance is key to understanding its meaning.224 223. Sauvé, supra note 220 at para 26, McLachlin CJ. 224. The ECtHR’s perception of its institutional position and the relationships it seeks to maintain with Member States within the Council of Europe play a key role in interpreting the ECtHR’s supposed “acceptance” of the UK’s aims: It is one thing to refrain from declaring the government aims illegitimate and another thing to say that it “accepts” them. See generally Hirst, supra note 214; Sauvé, supra note 220. Cf Plaxton & Hardy, supra note 218 at 119 (arguing that the court “display[s] a staunch deference to the state’s own appreciation of the limits on the right”). Moreover, in light of the fact that later in the same judgment the ECtHR declared that the UK law was in breach of the Convention, it seems more appropriate to consider the ECtHR’s maneuvers here as a tactical retreat. Be that as it may, the ECtHR’s judgment demands a dialogical reading beyond the purely propositional or declaratory. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 471 2 IV. THE LANGUAGE OF RULES AND LEGAL POSITIVISM In the first half of this article, I have developed a Bakhtinian understanding of language. The sections that follow in the second half will pit this dialogical view directly against the views of legal positivists, who think paradigmatically in terms of legal propositions or sentences complete on their own. Take the famous “No vehicles in the park” rule put forward by H.L.A. Hart to illustrate his distinction between easy and hard cases, and the distinction between the core and penumbra of rules. Hart conceived of a core of settled and uncontroversial meanings of the term “vehicle” such that the application of the rule was clear, and another zone of penumbra in which the answer was uncertain (e.g., is an electrically-propelled car a vehicle?).225 Lon Fuller famously criticized Hart’s underlying semantic theory, in which meaning consisted in finding the dictionary meaning or common use of the word “vehicle,” whereas Fuller presents an alternative scenario of a statue of a vehicle, which would surely not trigger the same prohibition. Fuller reminds us that in the task of interpretation “we commonly have to assign meaning not to a single word (e.g., vehicle) but to a sentence, paragraph, or whole page or more of text” further arguing that a paragraph does not have a “standard instance” or “core meaning” that remains constant whatever the context.226 Legal positivists after Hart have acknowledged these criticisms and have suggested extending interpretation to pragmatic considerations.227 But the problem with such analyses runs deeper than pragmatics: It is in the underlying view of language which affects even the selection of examples like “No vehicles in the park.”228 225. The example was first raised by Hart in 1958. See “Positivism and the Separation of Law and Morals” (1958) 71 Harv L Rev 593 at 611. It was famously debated by Lon Fuller. See “Positivism and Fidelity to Law: A Reply to Professor Hart” (1958) 71 Harv L Rev 630 at 662-63. 226. Ibid at 663. 227. Kramer, supra note 176. Some positivist authors still defend the semantic theory of meaning. See e.g. Schauer, supra note 138. Schauer argues that linguistic symbols carry meaning independent of the communicative goals of the users of those symbols on particular occasions. 228. Legal positivists have a fondness for certain kind of examples: vehicles, cricket, chess—but not war, poverty, politics, or religion. Examples are never neutral (I thank Peter Goodrich for reminding me of this crucial point). 472 (2022) 59 OSGOODE HALL LAW JOURNAL Against the grain of the positivist view of language (and law), Roderick MacDonald and Jason MacLean imagine an alternative sign, “No toilets in the park,” which shows no grammatical or syntactical difference with Hart’s imagined rule.229 They envision this sign posted at the entrance of a public campground, accompanied by a pictogram with a red circle and diagonal line superimposed upon a water faucet issuing drips into a cup. MacDonald and MacLean suggest that we tend to interpret “no vehicles in the park” as a sign prohibiting certain action, whereas the sign “no toilets in the park” would normally be read as offering information (“The water is not potable”) or perhaps expressing warning (“You Are Advised Not to Drink the Water”). In other words, these two signs, which are linguistically equivalent, lead to very different normative conclusions. This means that “[n]o theory of legal interpretation that rests only upon words, grammar, and syntax …can alone answer why…‘No Toilets in the Park’ will typically be understood as descriptive and ‘No Vehicles in the Park’ will usually be read as prohibitive.”230 Could Bakhtin help us suggest that these two signs have different intonations? The interpretation of the sign prohibiting vehicles in the park may further vary depending on the park and the kind of activities imagined there: We can think of recreation areas for family picnics, where the noise and disturbance occasioned by vehicles may frustrate the experience; alternatively, in the context of a safari park where animals often roam freely, the prohibition of vehicles may have a more ominous interpretation (is a wild lion attacking vehicles?). Signs can also be read differently depending on location: In a city park, a sign like the one imagined by MacDonald and McLean about the lack of drinkable water announces a mere inconvenience, but in places where water is a precious commodity, the sign may well remind us of our environmental responsibilities and raise questions of political economy around access, privilege, and inequality. The meaning of the sign is connected with these imagined scenes: the circumstances, types of activities, underlying rationales, possible exceptions,231 and multiple other factors that go “without saying” and remain unsaid.232 All this is to say that a lot goes into the process of making sense of a seemingly simple sign. As Jeremy Webber has 229. Roderick A MacDonald & Jason MacLean, “No Toilets in Park” (2005) 50 McGill LJ 721 at 723-27. 230. Ibid at 724. 231. See Steven Winter, “Frame Semantics and the ‘Internal Point of View’” in Freeman & Smith, eds, Law and Language (Oxford University Press, 2013). 232. Marianne Constable & Linda Ross Meyer, “No rule for the application of rules” in Baudouin Dupret, Julie Colemans & Max Travers, eds, Legal Rules in Practice: In the Midst of Law’s Life (Routledge, 2021) 29 at 35. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 473 argued, no rule can be understood without the set of practices and background experiences that inform its interpretation, and these are often non-propositional in kind.233 Because no rule is self-executing, the rule printed on a sign cannot determine its own interpretation and application: All additional information is provided by users of the park in their interactions with the posted sign.234 To be sure, much has been written on the open texture of language, the vagueness of rules, and the attempts to curb them since Hart.235 From an opposite, critical perspective, plenty has been made of the indeterminacy of language and the inevitable gap between rule and interpretation.236 Rather than parse through the various arguments and theories, I want to bring the debate to the latest linguistic trend for statutory interpretation, called corpus linguistics. V. CORPUS LINGUISTICS AND STATUTORY INTERPRETATION Corpus linguistics is the study of language-use in large digital collections of naturally occurring language, called “corpora” (plural of corpus). These databases can contain entire libraries, vast collections of historical materials, newspaper articles and magazines, transcripts of conversations, personal letters, et cetera.237 233. See “The Grammar of Customary Law” (2009) 54 McGill LJ 579 at 588-89. As Webber writes, “[S]tatements of norms…are always approximations, distillations, interpretations, that are perennially subject to further evaluation and refinement as a result of experience” (ibid at 588). 234. According to MacDonald & MacLean, “the interpreter is not merely an exegete,” that is, someone who must reproduce the meaning of the rule, but “an agent engaged in constituting both the meaning of the sign and his or her own relationship to it.” Supra note 229 at 725-26. 235. There have been some important attempts in the Anglo-American world. See e.g. Bix, supra note 9; Schauer, supra note 138; Timothy Endicott, Vagueness in Law (Oxford University Press, 2000). See also Asgeirsson, supra note 168. 236. See e.g. Duncan Kennedy, Critique of Adjudication: Fin de Siècle (Harvard University Press, 1998); Stanley Fish, Is There a Text in the Class? The Authority of Interpretive Communities (Harvard University Press, 1982). On differànce, see Jacques Derrida, Writing and Difference (Chicago University Press, 1978) [Derrida, Writing]. For Derrida’s understanding of language as trace, opposed to its “presence” in consciousness, see also Jacques Derrida, Voice and Phenomenon: Introduction to the Problem of the Sign in Husserl’s Phenomenology (Northwestern University Press, 2011); Jacques Derrida, Of Grammatology, 40th ed, translated by G Spivak (Johns Hopkins University, 2016). 237. For a good short introduction into corpus linguistics, see Clark Cunningham & Jesse Egbert, “Analyzing legal discourse in the United States” in Eric Friginal & Jack Hardy, eds, Routledge Handbook of Corpus Approaches to Discourse Analysis (Routledge, 2020) 462. 474 (2022) 59 OSGOODE HALL LAW JOURNAL Animated by a positivist mindset and computer-generated search engines, corpus linguistics seeks to ascertain the ordinary, typical, or common meaning of a given legal provision (i.e., a piece of legislation, a statute, the Constitution) by analyzing how the terms are reflected in the corpora.238 Aiming at “objectivity and predictability,”239 corpus linguistics thus hopes to bridge the divide between textualists and purposivists by means of systematic and empirical scrutiny.240 One of the methods employed by corpus linguistics is to survey actual speakers about which of the possible meanings of a term is the most prototypical (i.e., the best and clearest) example of its kind. For example, a pioneering study in a legal setting was done by Clark Cunningham in collaboration with a group of linguists in order to ascertain whether the term “enterprise” used in the United States’ Racketeer Influenced and Corrupt Organizations Act (RICO) denoted mostly economic or non-economic activity.241 Investigations such as these aim to set the boundaries of words by capturing both a term’s definitional content or intension (e.g., the specific qualities or attributes that make a sandwich a sandwich), as well as the term’s extension (e.g., the range of cases to which the name “sandwich” applies). Rather than relying on the more or less intuitive assessment of native speakers about the ordinary meaning of words, another method looks at their frequency in the corpora. According to this view, “the common usage of a given term in a given context is an empirical matter that may be quantified through corpus-based methodologies.”242 Notably, Thomas Lee and Stephen Mouritsen, prominent advocates of this methodology, have applied it to the Hartian rule prohibiting vehicles in the park.243 Lee and Mouritsen test the disagreement in the corpora and reach the conclusion that the most common use of the word ”vehicle” includes automobiles, ambulances, and golf carts, but not bicycles or airplanes.244 Critics of corpus linguistics have pointed out—and proponents do not deny—that the methodology is not exempt from selection and confirmation 238. See e.g. Stefan Gries & Brian Slocum, “Ordinary Meaning and Corpus Linguistics” (2017) 2017 BYUL Rev 1417. 239. Stephen Mouritsen, “Hard Cases and Hard Data: Assessing Corpus Linguistics as an Empirical Path to Plain Meaning” (2011) 13 Colum Sci & Tech L Rev 156 at 180. 240. See Cunningham & Egbert, supra note 237. 241. 18 USC, c 96 (1970); Clark Cunningham et al, “Plain Meaning and Hard Cases” (1994) 103 Yale LJ 1561. 242. Mouritsen, supra note 239 at 162. 243. Thomas Lee & Stephen Mouritsen, “Judging Ordinary Meaning” (2018) 127 Yale LJ 788. 244. Ibid at 836-44, 859-62. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 475 biases.245 Further, “not every case of legal interpretation presents a neat, binary question of lexical ambiguity,” such as “is a bicycle a vehicle or not,” and “[i]n some cases the statutory language may be hopelessly vague.”246 Additionally, the corpus may not capture the specificity and institutional context of legal language (e.g., the value of legal precedents and specific terminologies used by courts).247 Lastly, the determination of a given term may depend on extra-linguistic circumstances that are hard to pin down in advance. For example, Lee and Mouritsen admit that the rule “no vehicles in the park” “may depend on the physical and spatial characteristics of the park itself.”248 But the problem with corpus linguistics runs deeper than limitations of method, databases, or ambiguous circumstances. The main objection is that corpus linguistics does not adequately represent what the interpretive task is all about. Consider again the “no vehicles in the park” scenario: If an ambulance driver has to cut through the park to save someone’s life, the question is not to decide whether the ambulance is or is not a “vehicle” but whether it is appropriate in that context for the driver to cut through the park, or alternatively, for a police officer to stop the ambulance and fine the driver. To think that the linguisticallycorrect interpretation of the rule consists in verifying the frequency in the use of a particular “vehicle” in the corpora is as misguided as seeking to determine whether a given punishment is “cruel” by measuring the convicted prisoner’s 245. For a criticism along these lines, see Carissa Byrne Hessick, “Corpus Linguistics and the Criminal Law” (2017) 6 BYUL Rev 1503. For an empirical criticism of the reliability of corpus methodology, see Kevin Tobia, “Testing Ordinary Meaning” (2020) 134 Harv L Rev 706. 246. Mouritsen, supra note 239 at 204. 247. See Evan Zoldan, “Corpus Linguistics and the Dream of Objectivity” (2020) 50 Seton Hall L Rev 401. For an insightful critique, see Anya Bernstein, “Legal Corpus Linguistics and the Half-Empirical Attitude” 106 Cornell L Rev 1397. See also Lawrence Solan, “Corpus Linguistics as a Method of Legal Interpretation: Some Progress, Some Questions” (2020) 33 Intl J Sem L 283. 248. Lee & Mouritsen, supra note 243 at 843. The admission seems damning, for if the rule prohibiting vehicles in the park ultimately depends on circumstances as random as the physical and spatial characteristics of the park, one wonders, what is to be gained with the corpus? 476 (2022) 59 OSGOODE HALL LAW JOURNAL pain.249 Computerized parsing will not help to answer the relevant questions, nor will it ever serve to replace the actual task of interpretation.250 The belief that words have natural boundaries that can be determined empirically by means of surveys, statistics, or computerized data relies on a positivist conception of language that is deeply problematic.251 While the methodologies of corpus linguistics appear more sophisticated than simply looking at dictionary entries for specific terms, the approach still replicates a “dictionary conception” of meaning that identifies potential, rather than actual, meaning—what Bakhtin called “abstract signification.”252 As we have seen, the meaning of an utterance is inextricably linked to its intonation, register, and genre; its prehistory and to the utterances it presupposes; the relative positions of the speaker and addressee; and the responsiveness it generates. These are constitutive parts of every utterance, without which the utterance cannot be understood. As a result, “ordinary meaning”—if such a thing can be said intelligibly—is not found in general language usage (no matter how wide the sample) but in what different speakers take it to mean in concrete, actual contexts of living dialogue, and these contexts are always in dispute. Meaning does not reside in words, but in the actual event of communication in which words (but not only words) are used.253 How would someone with a Bakhtinian understanding of language approach the activity of legal interpretation? More specifically, what would it mean to read legal provisions not as sentences, but as utterances? First and foremost, we must 249. Not that the prisoner’s pain is irrelevant—far from it, the prisoner’s pain helps to explain why the prohibition of cruel punishment matters—but the task before the judge is not to empirically verify the pain of the prisoner. 250. Clark Cunningham & Charles Fillmore, “Using Common Sense: A Linguistic Perspective on Judicial Interpretations of ‘Use a Firearm’” (1995) 73 Wash L Rev 1159. See also Bernstein, supra note 247. Bernstein argues that “turning to an empirical method can distract legal interpreters from the inquiry they are charged with resolving” (ibid at 22). 251. This is what Neal Goldfarb calls the “slice-and-dice” approach to meaning. See “A Lawyer’s Introduction to Meaning in the Framework of Corpus Linguistics” (2017) 2017 BYUL Rev 1359 at 1390. As Bernstein argues, “legal corpus work routinely fails to even discuss issues of speaker, audience, and genre” (supra note 247 at 39-40). 252. Bakhtin, “Discourse,” supra note 1. See Patrick Hanks “Do Word Meanings Exist?” (2000) 34 Computers & Humanities 205 at 210-11. As lexicographer (and former chief editor of Current English Dictionaries at Oxford University Press) Patrick Hanks writes: [M]eanings are events, not entities….It is a convenient shorthand to talk about ‘the meanings of words in a dictionary’, but strictly speaking these are not meanings at all. Rather, they are ‘meaning potentials’—potential contributions to the meanings of texts and conversations in which the words are used, and activated by the speaker who uses them. 253. See also Bernstein, supra note 247 at 49 (arguing that “[m]eaning is not a fact, [but]…a social activity”). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 477 accept that as an utterance, no legal provision can be individually plucked from the web of prior texts and responses that weave together a rich “culture of argument.”254 From a Bakhtinian standpoint, the object of interpretation is not the legal provision, proposition, or rule considered in the abstract, but rather the language as fleshed out in the particular scene, as we saw in the example of the ”No water” sign in the park.255 To say it differently, legal interpretation is not to attribute meaning to a piece of text but to the situations in which the language of the law is used. In a scene of legal interpretation, meaning is not a property of words (or sentences) but rather of utterances in which words are put to work. Legal interpreters speak from particular positions and bring something of their own to the scene: They augment it, enrich it, push it forward—or else subtract from it, distinguish it, shore it to their own advantage. Because “understanding comes to fruition only in the response,” legal interpretation is not duplicative or reiterative but responsive and generative.256 Any account of interpretation must take these aspects of living language into account, which I illustrate with examples from a well-known series of cases decided by the Supreme Court of the United States about the expression “using or carrying a firearm” that have been the subject of much academic debate in corpus linguistics.257 1. Legal utterances cannot be stripped from the rich background of prior texts and set up against a neutral language. Corpus linguists seek to establish the ordinary or common meaning of a statutory term, but in order to do so they must abstract the legal provision from its actual context. For example, in Smith v. United States (“Smith”),258 the defendant was arrested after trying to exchange an automatic weapon for two ounces of cocaine from an undercover officer. John Smith was sentenced to a mandatory 30-year enhanced sentence, pursuant to a 254. See James Boyd White, “Constituting a Culture of Argument: The Possibilities of American Law” in When Words Lose Their Meaning: Constitutions and Reconstitutions of Language, Character, and Community (University of Chicago Press, 1984) 231. 255. For the concept of the legal scene, see Julen Etxabe, “Jacques Rancière and the Dramaturgy of Law” in Mónica López Lerma & Julen Etxabe, eds, Rancière and Law (Routledge, 2018) 17. 256. Bakhtin, “Discourse,” supra note 1 at 282. 257. Smith v United States, 508 US 223 (1993) [Smith]; Bailey v United States, 508 US 223 (1995) [Bailey]; Muscarello v United States, 508 US 223 (1998) [Muscarello]. For the first corpus linguistic analysis, see Cunningham & Fillmore, supra note 250. For more recent analyses, see Gries & Slocum, supra note 238; Goldfarb, supra note 251; Lee & Mouritsen, supra note 243. 258. Supra note 257. 478 (2022) 59 OSGOODE HALL LAW JOURNAL federal provision criminalizing the use of guns in drug-related crimes.259 The issue at hand was whether exchanging a firearm constitutes “use of a firearm,” but the answer cannot be divorced from the long history of drug-related gun violence in the United States. Thus, the majority of the US Court in Smith mentioned that in 1989, 56 per cent of all murders in New York City were drug-related, and in Washington D.C. the figure was as high as 80 per cent.260 Furthermore, the majority placed special emphasis on the particulars of the gun in question, describing it as “a favorite among criminals” and pointing out that it had been modified to “fire more than 1,000 rounds per minute.”261 The petitioner sought to avoid the enhanced penalty of the statute by arguing that exchanging the gun for drugs did not constitute “using a firearm.” But when placed alongside its social dangerousness, the distinction that the petitioner wanted to draw between using the gun “as a weapon” and using it “as an item of barter” was ruled as “metaphysical,”262 given that in both instances, the possibility of drug-related gun violence increases all the same.263 The question for the majority was not whether ordinary language allows this distinction but whether it “made sense” in the situation to distinguish in this way. For our purposes here, the issue is not whether the majority was right or wrong in its conclusion but that the US Court did not and could not test the legal provision against an allegedly neutral language unaffected by prior cultural texts and narratives. 2. A legal utterance is part of a wider web of utterances to which it is related. A legal provision, no matter how novel and revolutionary in orientation, comes already languaged,264 and it enters a universe of prior languages, previously decided cases, doctrinal elaborations, nested narratives, and institutional networks. Corpus linguistics may seek fresh meanings in computer-generated searches,265 but new words are enmeshed in older interpretations that oblige any interpreter to account 259. Federal Criminal Code, USC § 924(c)(1)(Title 18). This provision imposes an enhanced mandatory sentence of five years for anyone who “uses or carries a firearm” during and in relation to a drug trafficking crime—and 30 extra years if the weapon is a machine gun or is fitted with a silencer. 260. Smith, supra note 257 at 240, O’Connor J. 261. Ibid at 225. 262. Ibid at 240. 263. As the majority writes, “It creates a grave possibility of violence and death in either capacity” (ibid). 264. Becker, supra note 2 at 287. 265. Mouritsen explains that when searching for the term “enterprise” in the Racketeer Influenced and Corrupt Organizations Act (RICO) statute, he excluded certain uses from the count “with an eye to treating the ordinary meaning of enterprise as a fresh linguistic question, rather than a long-decided legal one.” Supra note 239 at 196. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 479 for tangled plots and winding trajectories. This is to say that the language of law has not only history but also memory.266 Legal participants contribute to the “threaded nature” of legal language by sequential co-weaving and intertwining utterances that unfold in time.267 The image of legal interpretation as a chain is associated with Ronald Dworkin’s ideal of integrity,268 but no legal tradition is perfectly continuous: They are ragged with moments of discontinuity, improvisation, and rupture.269 Every tradition contains powerful jurisgenetic moments that cannot be understood within a model of linguistic or normative coherence.270 Indeed, jurisprudential traditions tend to prize consistency, but when the fiction of continuity cannot be sustained, new pathways can be forged. Thus, deviating from their conclusion in Smith, the Supreme Court decided only two years later in Bailey v. United States (“Bailey”) that “using a firearm” requires “active employment” of it.271 And though Justice O’Connor, who wrote both judgements, stated that Bailey “is not inconsistent with Smith,”272 her pronouncement reads more like an attempt not to dispel the myth of continuity of precedent than as its actual manifestation. 3. Legal utterances cannot be ascertained independently from the relative position of the speakers. In the propositional conception of language, the position of the speakers is irrelevant to the content of their speech and should not affect their interpretation. From a Bakhtinian perspective, however, the meaning of the law governing use of firearms depends on the institutional position and role of the interpreter. A prosecutor may reasonably indict someone for trying to hide a gun in the trunk of a car—which a defender will likely oppose on the grounds that 266. On the memory of language, see Boris Gasparov, Speech, Memory, and Meaning: Intertextuality in Everyday Language (De Gruyter, 2010). 267. While generally associated with the common law, civil law and non-Western traditions also show a “threaded nature,” as shown by Baudouin Dupret and Jean-Noël Ferrié in their ethnomethodological study of the interplay between secular and Shari’a law in Egypt. See “The Practical Grammar of Law and its Relation to Time” in Baudouin Dupret, Michael Lynch & Tim Berard, eds, Law at Work: Studies in Legal Ethnomethods (Oxford University Press, 2015) 27. 268. Law’s Empire (Harvard University Press, 1986). 269. See Alexandre Lefebvre, The Image of Law: Deleuze, Bergson, Spinoza (Stanford University Press, 2008). See also Sara Ramshaw, Justice as Improvisation: The Law of the Extempore (Routledge, 2013). 270. See Julen Etxabe, See Julen Etxabe, The Experience of Tragic Judgment (Routledge, 2013) [Etxabe, Tragic Judgment]. Cf Cunningham and Fillmore, who analyze the legislation history—i.e., the successive phases of legislative changes and reform—“as if it were a coherent discourse over time.” Supra note 250 at 1163. 271. Bailey, supra note 257 at 143. 272. Ibid at 148. 480 (2022) 59 OSGOODE HALL LAW JOURNAL the gun could hardly have been “used” from there.273 For justices of the Supreme Court who must consider the statute as well as the constitutional rights of the defendant, the relevant words may appear in yet a different light. As the US Court unanimously stated in Bailey, the words of the statute “cannot support the extended applications that prosecutors have sometimes placed on it, in order to penalize drug-trafficking offenders for firearms possession.”274 In other words, in interpreting the provision, the Court was not acting as a “reasonable hearer of speech,” which is the role granted to them by positivist linguistics.275 Instead, the US Court interpreted the provision from its concrete institutional position and in light of its responsibilities as a Supreme Court. 4. Legal interpretation is part of the legal scene in which it appears. Neither a legal provision nor its interpretation is separable from the situation in which it is invoked and performed. The interpreter reads the legal utterance against the background of the particular scene in which the legal provision is embedded. The legal provision cannot be tested simply by analyzing different uses of the terms in the linguistic database, no matter how large or extensive, because databases do not contain the particular event at present—whether the scene involves a gun being found in the trunk of a car during a traffic stop (as in Bailey) or an individual trying to exchange a MAC-10 for two ounces of cocaine (as in Smith); the exact same provision may appear differently.276 Among other things, the scene is constructed (or staged) by the encounter between the different actors, the language of the speakers, the prior texts they inherit, their evaluative orientations and background assumptions, the setting in which the dispute arises, the audiences and their responsiveness, the imagined consequences to follow, et cetera. Yet, the legal scene is not uncontroversially shared between the actors. For example, the majority in Smith refused to draw a distinction between using a gun as a weapon and using it as a means of exchange even though it meant the defendant would face an additional sentence of thirty years. As Justice Ginsburg 273. Ibid. In one of the two petitions consolidated in Bailey, the police found a pistol in the defendant’s trunk. 274. Ibid at 150, O’Connor J. 275. This position can also be mediated by several legal constructs (or, rather, by the Supreme Court’s interpretation of them), such as the “average person on the street,” the “reasonable member of Congress,” the “normal speaker of English,” the “reasonable hearer.” For some of these categories, see Slocum, supra note 174 at 21-24. 276. The scene is not merely a cognitive process in the mind of the interpreter. Nor does it allude just to the need of creating coherent narratives of events. See e.g. Bernard S Jackson, Law, Fact and Narrative Coherence (Deborah Charles, 1988). Rather, the scene encapsulates the interpreter and brings the interpreter in. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 481 pointed out in her dissent in the subsequent case of Muscarello v. United States, the scene ought to be enlarged to incorporate fundamental principles of criminal law and punishment, due to the “blunt” mandatory enhanced sentence and the “instinctive distaste against men languishing in prison unless the lawmaker has clearly said they should.”277 Moreover, a scene may be complicated by subtext—silent scripts that can suddenly be given expression. One important undercurrent to any American discussion about guns is the Constitutional mythos that enshrines the right to possess firearms. In Smith, this counternarrative did not come to the surface, maybe because it is difficult to imagine the mythos working on behalf of a drug-trafficking defendant who was apprehended with two machine guns, a loaded .45 pistol, a .22 calibre pistol with a scope and homemade silencer in his van, and a loaded 9-millimetre handgun in his waistband. However, in Muscarello, where the defendant was carrying a gun in the glove compartment of his car, the social imaginary was brought explicitly into the open, on the grounds that the Second Amendment frames “carrying a gun” as bearing it on the person.278 5. The interpretation of legal utterances is internally dialogized. Those who believe that they rely on ordinary meaning often presuppose that native speakers share the same linguistic assumptions and apply the same standard to legislative texts.279 Not only do these assumptions ignore language differentiation between speakers and social groups, but even within the same epistemic and interpretive communities, different members of the legal profession (different judges, prosecutors, defence attorneys, et cetera) need not share common understandings of particular terms. Every legal provision enters a dialogically-agitated universe in which meanings are in tension. Therefore, it is a mistake to assume, as Justice Breyer did in Muscarello, that words have an ordinary meaning that remains constant throughout every instance of application, as if carrying “a gun, a suitcase, or a banana” made no difference for the purposes of the provision.280 Surely, carrying a gun and carrying a banana do not have the same legal implications 277. Supra note 257 at 150. Ginsburg J, dissenting, citing United States v Bass, 404 US 336 (1971). 278. Supra note 257 at 143, Ginsburg J, dissenting. 279. Slocum, supra note 174 at 14. See also Schauer, supra note 138. 280. Muscarello, supra note 257. Justice Breyer argued that “the word ‘carry’ in its ordinary sense… keeps the same meaning whether one carries a gun, a suitcase, or a banana” (ibid at 131 [emphasis added]). 482 (2022) 59 OSGOODE HALL LAW JOURNAL and no abstract meaning of “carrying” that can encompass both actions.281 If any such thing as an “ordinary” or “common” meaning is said to exist, this is to be defended against alternative considerations of what “common” and “ordinary” might mean in the disputed context. 6. Legal interpretation is not reproductive, but dialogic. Corpus linguistics seeks to establish the ordinary meaning of legal provisions, as if meaning were a matter of empirical verification rather than something to be appraised in one context and refurbished in another. In claiming that the meaning of statutes is asserted at the moment of promulgation, the supposition is that this meaning must be reproduced mimetically at the time of interpretation, in perfect correspondence between the two. In this account, the interpreter aspires to be a mere conduit, adding nothing that is not there already.282 In other words, the ideal is for interpreters to become “transparent” and let the text speak on its own.283 But however popular the image of a text that “speaks on its own” may be, as an account of the hermeneutic process it is far from adequate. As Hans-Georg Gadamer elucidated, the ideal of a transparent interpreter who can become one with the original text or author denies the process of historical mediation. Between the moment of promulgation and the moment of interpretation, there is a temporal gap that is never completely sutured by an act of “application.”284 On the other hand, with our slightly modified Bakhtinian template, we can imagine not an individual interpreter faced with a legal text that is already replete, but speakers engaged with other speakers in dialogical encounters (past, present, and future). The interpreter appraises the legal words that come before, finds a place for them in the interpreter’s own discourse, anticipates a possible comeback, and invites a form of criticism for the future. This is why interpretation has no tautology or duplication. A legislator may write the words in a statute, but when interpreting how the provision demands to be responded to, the interpreter 281. Ibid. Justice Breyer seeks to isolate the word “carry” independently of any context, as if the word existed in a Platonic world of pure concepts. Surely, carrying a gun or carrying a banana have different normative consequences and are therefore not the same action: There is a reason why the statute talks about guns and not bananas. 282. See e.g. William Eskridge Jr, “The New Textualism” (1990) 37 UCLA L Rev 621. This approach is often associated with textualism, brought to prominence by Justice Scalia. New textualism is, however, not the province of conservative scholars; it has been repurposed by liberal constitutional scholars like Akhil Amar, Lawrence Lessig, and Jack Balkin. See e.g. James Ryan, “Laying Claim to the Constitution: The Promise of New Textualism” (2011) 97 Va L Rev 1523. 283. On this idea of transparency, see Paul Kahn, Making the Case: The Art of the Judicial Opinion (Yale University Press, 2016). 284. Hans-Georg Gadamer, Truth and Method, 2nd ed (Continuum, 1975). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 483 augments and adds depth to it from a different context or situation, pushing it forward in new directions. The task of legal interpretation is hence not to reproduce the meaning of the legal text, but to answer its call. As a case in point, when interpreting the words “to use a firearm,” the US Supreme Court in Bailey required the prosecution to “show an active employment of the firearm.”285 In no way can this interpretation be described as merely the linguistic elucidation of ordinary meaning. Rather, the US Supreme Court responded to the question of interpretation by embedding the provision within foundational principles of criminal law (the presumption of innocence, in dubio pro reo, and non-retroactivity); by referring to the rules of evidence and the burden of proof; by expressing institutional concerns about the role of courts to review governmental action; and by reaffirming democratic principles about transparency, the need to justify the coercive authority of the state and the limits of the criminal law. The US Supreme Court’s judgement thereby demonstrates the multiplanar and thick dialogic process that legal interpretation actually entails. VI. THE ALLEGED NEUTRALITY OF DESCRIPTIVE LEGAL POSITIVISM The rule-based semantic approach to legal interpretation and corpus linguistics are both reflections of a foundational tenet of legal positivism, namely, that it is possible to have a general, descriptive, and morally neutral theory of law. As Hart put it, this theory “does not seek to justify or commend on moral or other grounds the forms and structures which appear in [the] general account of law.”286 Legal positivism’s self-presentation of neutrality notwithstanding, Hart’s own theorization of law as the union of primary and secondary rules entails many evaluative judgments, as highlighted by Stephen R. Perry in an insightful article.287 Hart’s theorization takes Western modern municipal systems as a starting point, establishing as outliers from the very beginning alternative manifestations of law, such as so-called “primitive” legal orders, international law, historical legal orders, and legal practices in culturally-diverse societies. In doing so, Hart made significant background assumptions. For example, he took for granted a fairly recent historical innovation, the modern state; he presupposed the existence of different kinds of rules, as well as institutions to adjudicate conflicts (courts) and to make and modify the rules (legislatures), thus replicating the structure 285. Bailey, supra note 257 at 143, O’Connor J. 286. The Concept of Law, supra note 173 at 240. 287. “Hart’s Methodological Positivism” (1998) 4 Leg Theory 427. 484 (2022) 59 OSGOODE HALL LAW JOURNAL of primary and secondary rules that he later theorized. In addition, among the conditions that Hart stipulated for the existence of a legal system, he prioritized the acceptance of secondary rules by officials, which entails implicit, though not always forthcoming, arguments of political philosophy. Hart further claimed that legal systems operating exclusively with primary rules are “defective” and cannot perform the tasks presumptively attributed to law, making evaluative claims “but about the very social practices he is studying.”288 Hence, while Hart claimed to be picking out “social practices as ‘law’ on the basis of purely factual, non-normative criteria,” he palpably did much more than that. In fact, when Hart presented his concept of law as the union of primary and secondary rules, he did not simply clarify or elucidate a concept of law that people already have—he effectively transformed it.289 Faced with this criticism, legal positivists have basically three options: to embrace normativism, to correct course, or to defend neutrality.290 The first response involves retracing the normative foundations of legal positivism by attending to figures like Jeremy Bentham who understood their theories to be engaged fundamentally in a normative endeavour. According to this reading, legal positivism is not inimical to, but rather predicated upon, moral and political ideals about the rule of law. This version of legal positivism, articulated by scholars like Tom Campbell and Jeremy Waldron, is called “ethical”291 or “normative,”292 because it adopts the view that the identification and application of law ought to be kept as separate as possible from moral judgments.293 The thesis is thus defended on political grounds rather than on conceptual ones, and more importantly, it still depends on the analytical possibility that the identification of law can be made without moral evaluation.294 (I will come back to ethical positivism in the conclusion) A second option for positivists is to view the slippage into normativity as a problem to be rectified and to attempt to correct course. They may accept that 288. Ibid at 438. 289. Ibid at 444. 290. These alternatives correspond to the ethical, socio-legal, and conceptual variants of legal positivism respectively. 291. See Tom Campbell, The Legal Theory of Ethical Positivism, reprinted ed (Routledge, 2016) [Campbell, Legal Theory]. 292. See Jeremy Waldron, “Normative (or Ethical) Positivism” in Jules Coleman, ed, Hart’s Postscript: Essays on the Postscript to ‘The Concept of Law’ (Oxford University Press, 2001) 411. 293. This is called the “prescriptive separation thesis” (Campbell, Legal Theory, supra note 291 at 3). See also Waldron, supra note 292 at 411. 294. Campbell, Legal Theory, supra note 291 at 2; Waldron, supra note 292 at 418. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 485 Hart failed to provide a wholly-descriptive and general concept of law, but at the core of such admissions is a continuing commitment to a “descriptive sociology,” which strives to eliminate all normative and ideological remnants from the theory. Brian Tamanaha, for example, has worked towards a bare or “thinned out” concept of law that attempts to avoid any normative criteria, functions, and requirements.295 The third possible response to the critique of legal positivism entails denying that Hart was engaged in any sort of normative endeavour altogether. (This is also compatible with Jules Coleman’s argument that Hart could have succeeded, even if he ultimately did not.296) According to this third position, legal positivism is essentially descriptive and morally neutral, because “the theory need not take a stance on any particular moral or political issues, nor is it committed to any moral or political evaluations.”297 The fact that some positivists espouse normative goals is irrelevant because “legal positivism makes a conceptual, or analytic claim about law, and that claim should be not confused with the programmatic or normative interests certain positivists…might have had.”298 This third approach admits that sometimes positivist theorists incorporate evaluations but, as Julie Dickson puts it, only in an “indirect” manner, which is very different from the kind of directly evaluative judgments of non-positivist theories (i.e., about the merits or demerits of given manifestations of law).299 My aim is not to challenge these jurisprudential positions directly but to lay bare the underlying conception of language that they share and that defines them as positivist. In a nutshell, legal positivism is based on a misguided assumption about what it means to use the language of description—and of 295. See Brian Tamanaha, “Socio-Legal Positivism and a General Jurisprudence” (2001) 21 Oxford J Leg Stud 1. Tamanaha seeks a “conventionalist” account, where law “lacks any inherently necessary qualities” and instead is “whatever we attach the label law to” (ibid at 15, 18). 296. See “Incorporationism, Conventionality, and the Practical Difference Thesis” (1998) 4 Leg Theory 381 at 395, n 26 (arguing that the criticisms purporting to show that Hart, contrary to his own reflections on the matter, did not engage in or faithfully execute a descriptivist methodology, fall short of undermining the very possibility of a descriptive jurisprudence). 297. Marmor, “Still Descriptive,” supra note 11 at 125. 298. Jules L Coleman, “Negative and Positive Positivism” (1982) 11 J Leg Stud 139 at 147. 299. On direct and indirect evaluations, see Julie Dickson, Evaluation, supra note 10 at 51-57. The distinction is less stable than Dickson makes it appear, for “what is driving [the decision about] the importance of certain feature of the law is…[that it] has a bearing upon, or is ultimately relevant to, answering directly evaluative questions” (ibid at 61). 486 (2022) 59 OSGOODE HALL LAW JOURNAL theorizing itself.300 Indeed, as Gerald Postema perceptively puts it, “[a]nalytical jurisprudence…mistakenly assumes that concepts we use can be divorced from the language of everyday life in which they function.”301 But “[c]onceptual analysis is not sharply distinct from the enterprise of gaining an understanding of the practices and forms of life in which the concepts have their life.”302 As defined by its practitioners, general jurisprudence is a reflective attempt to understand our most general beliefs concerning legal practice, including beliefs and attitudes that form our legal sensibilities.303 “Jurisprudential theory, then, even when it appears to be engaged in conceptual analysis, is focused on the task of giving an account of legal institutions, and the practice and ‘sensibility’ that breathe life into them.”304 This can never be a matter of simply pointing at social facts, but is rather one of characterization. Drawing on Charles Taylor, Postema has suggested simple descriptions can be accurate or inaccurate, and their inaccuracy can be shown by pointing to new or overlooked facts or evidence, but when characterizations fail they distort the reality they seek to interpret. And, after a point, no mere showing, or marshalling of more evidence, will settle a dispute regarding the truth of the characterization. That question will turn on the strength and plausibility of the sense, point, or meaning attributed to those facts. Thus, because they do not merely ascribe properties to objects, but instruct us about how to think about them, characterizations do not leave the phenomena unchanged.305 Postema’s observations pave the way for several important considerations. To begin, legal theorizing in general, and conceptual work in particular, do not operate on a different plane than the rest of language, and are subject to the same vagaries and complications. It may be possible for legal positivists to offer schematic definitions—e.g., law is the union of primary and secondary rules, or law is a coercive normative order—but these statements capture neither the complexity nor the depth, let alone the beliefs and attitudes, of those who live by law (or under it).306 More significantly for my argument here, insofar as the goal of jurisprudence is to “breathe life” into legal practices and institutions, pithy and 300. For the imaginative character of this endeavour, see William MacNeil, Novel Judgements: Legal Theory as Fiction (Routledge, 2012). 301. Bentham and the Common Law Tradition, 2nd ed (Oxford University Press, 2019) at 324. 302. Ibid at 324-5. 303. Ibid at 325. 304. Ibid. 305. Ibid at 326. 306. According to Postema, such abstract concepts and definitions “may capture only a small part of the body of beliefs and attitude which constitute the practice and shape the concept” (ibid at 324). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 487 reductive statements such as “law is a coercive normative order” do not qualify as what we ordinarily understand to be offering a good “account.”307 This is because, as explained by Postema, a theoretical account does not merely ascribe properties to objects but instructs us about how to think about them. In other words, there is no neutral referent of law against which we may assess the accuracy of our descriptions, and we sooner speak of accounts that are insightful, compelling, significant, illuminating, thought-provoking, nuanced, and so forth. To do legal-theoretical work is an active exercise of selecting, filtering, tailoring and refining, imagining scenarios, testing opposing views, redescribing outdated language, proposing new ways of looking at it, et cetera. All of these tasks are evaluative from beginning to end. As explained by Webber, “[T]he very language that participants use to conceptualize and analyze—the very concepts they employ to conceive of normative challenges— have particular normative dispositions inscribed in them.”308 In fact, The understanding of what constitutes a fact and what constitutes a law, who are legally relevant agents, what are their intrinsic relations, how are they located within the natural world, what types of relations are written into that world, perhaps even what counts as “order”—are all structured, at least in some considerable measure, by presuppositions that...are built into the terms that people bring to the very effort to coordinate their actions.309 We have seen that the second variant of legal positivism attempts to address these concerns by dropping all normative criteria. However, even deliberately bare concepts of law are not outside language and cannot be stripped of normative presuppositions all the way down. Thus, Tamanaha’s socio-legal positivism is “unflinchingly conventional,” except for the very categories of analysis, which “are designed to meet the purposes of the theoretician who constructs them.”310 This problem becomes more acute the farther we move from our own coordinates. For example, when approaching the law of distant times and places—e.g., ancient Mesopotamia, classical Rome, or the Middle Ages, let alone legal practices such as oath-taking, the torture of slaves, and the criminal trials of animals311—we are confronted by different cosmologies, rationalities, and rules of causation; different considerations of the natural and social worlds; and 307. 308. 309. 310. 311. Ibid at 325. Supra note 233 at 605. Ibid at 604. Supra note 295 at 31. See William Ewald, “Comparative Jurisprudence (I): What was it Like to Try a Rat?” (1995) 143 U Pa L Rev 1889. 488 (2022) 59 OSGOODE HALL LAW JOURNAL different symbolizations of law’s force, shape, and authority.312 To try to bring all these under the same legal categories is inevitably to miss or distort some of them. In modern times, similar objections have been raised by anthropologists, legal pluralists, Indigenous legal scholars, postcolonial writers, feminists, and critical legal scholars against the hegemonic ambitions of Western law, its formal and written scripts, its abstract forms of reasoning, and the dismissal of mythical, cultural, and aesthetic dimensions. These dilemmas cannot be resolved by moving the analysis to a still-higher plane of abstraction where our concepts would presumably be free from evaluations, as Tamanaha intended, for here’s the paradox: The more we seek to abstract our language in order to avoid normative assumptions, the less detail our analysis will be able to provide and vice versa; the more and deeper our language wants to penetrate into legal practices, the more it will be embedded in contested vocabularies and assumptions. In sum, jurisprudential work remains inextricably tied up with our language, and this language is inevitably loaded.313 As we have seen, a third variant of positivism responds to this challenge by doubling down on moral neutrality and distinguishing between different types of evaluation. This view accepts that legal positivists bring assumptions to their object of study, and that their theories depend on cultural presuppositions and background experiences to get off the ground; but this is not to say that these are evaluative “in the relevant sense.”314 Thus, in suggesting that the sharpness of 312. See Henry S Maine, Ancient Law: Its Connection with the Early History of Society and its Relation to Modern Ideas, 10th ed (Henry Holt, 1906) at 6-7. As Maine critically observed in 1861, “the farther we penetrate into the primitive history in thought, the farther we find ourselves from a conception of law which at all resembles a compound of the elements which [legal positivists] determined” (ibid at 7). See also Clifford Geertz, Local Knowledge: Further Essays in Interpretive Anthropology (Basic Books, 1983). For Geertz, this suggests that “[l] ike sailing, gardening, politics, and poetry, law and ethnography are crafts of place: they work by the light of local knowledge” (ibid at 167). For the methodological challenges of understanding law in ancient societies, see Julen Etxabe, “Introduction: Writing a Cultural History of Law in Antiquity” in Julen Etxabe, ed, Cultural History of Law in Antiquity (Bloomsbury, 2019) at 1. 313. The critical question for any theory of law, then, is not whether it is evaluative or not, but what exactly does it enable us to see? And what does it obscure? What ontologies (textual, material, sensorial) are taken as fundamental, and which are downplayed? What epistemologies and forms of reasoning (analytical-cognitive, experiential, metaphorical) are considered, and which are excluded? What affective and aesthetic registers are highlighted? What chronotopes and spatial configurations are imagined—or ignored? 314. By morally neutral, Marmor means that “the theory need not take a stance on any particular moral or political issues, nor is it committed to any moral or political evaluations” (“Still Descriptive,” supra note 11 at 125). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 489 the knife makes it more useful, we are not making an evaluation of its various possible uses (from cutting bread to killing a person). This is essentially what Hart also said: “[L]aw is a tool, and…tools can be used for good or bad purposes alike.”315 For example, Marmor makes a distinction between grasping a value and having an evaluative attitude towards it, so that we should be able to understand the glorification of Catholicism in counter-Reformation Baroque architecture without having to admire either Catholicism or Baroque architecture.316 I will address this example more fully in the following part, when clarifying the confused idea of “detached normative statements,” but one phenomenological lesson we have learned from Bakhtin is that only by valuing something—i.e., showing concern for and answering it—do we fully experience (and understand) it.317 In other words, it is not that we value what we perceive, but that we perceive what we value. Accordingly, we do not first identify the law and then evaluate it, rather we do both at once. Positivists are keen on severing the descriptive moment, when law’s characteristics are selected from the (relevant or directly) evaluative moment, or when judgments on the merits are performed. In this vein, positivists suggest that being able to describe law as a tool that can be used indistinctly for good and evil is not to evaluate whether the instrument thus used is good or bad. My point is simple: By the time we decide that law is “a tool,” all the “relevant judgments” have already taken place. From this perspective, the objection is that legal positivism is like the proverbial hammer to which everything starts to look like a nail. Indeed, to qualify law as an instrument that can be indistinctly manipulated, perhaps even completely mastered, sits unwell with those legal traditions and forms of life for whom law is not—and cannot in a meaningful way be—thus viewed. Therefore, when saying that the descriptive task can be done while avoiding evaluations “in the relevant sense,” one suspects that so-called direct or relevant judgments have not been avoided but simply pushed back—deferred, displaced, repressed, or elevated to a higher plane of abstraction. 315. Ibid at 143. 316. Ibid at 147. 317. Bakhtin, Act, supra note 142. 490 (2022) 59 OSGOODE HALL LAW JOURNAL VII. DEBUNKING “DETACHED NORMATIVE STATEMENTS” If legal theory is a deeply evaluative enterprise, the question arises now as to whether it is possible to state “what the law is” in a value-neutral fashion as defended by legal positivism. Based on a reading of Hans Kelsen, Joseph Raz has answered affirmatively by way of what he calls “detached normative statements,” which have been widely accepted in the positivist literature as the very register in which legal positivism proceeds.318 According to Raz, detached normative statements are a distinct type of legal utterance, situated between the statements of social scientists who describe people’s beliefs, habits, practices, and attitudes (what Hart called the “external point of view”) and the internal statements of those who accept (i.e., endorse) the values of the various laws and regulations being described.319 Detached legal statements are statements of law or legal obligation from the point of view of one who accepts the laws in question as valid, but without committing the speaker to them.320 Raz offers the example of a Catholic who is an expert on Rabbinical law and offers advice to a relatively uninformed orthodox Jew about what he is to do according to Jewish law.321 Similarly, an anarchist law professor might be able to explain the content of positive law as a system of valid norms without having to approve of them or their moral authority.322 According to Raz, much of the legal discourse is of this kind, for instance, the exposition of law professors before their students or the advice of lawyers to their clients. As Raz has explained, detached statements assert what the law is from the point of view of “the legal man,” a hypothetical person who accepts all and only the norms of the legal 318. Raz attributes this view to Kelsen, but Raz’s reading of Kelsen is more a development of his own project. On Kelsen’s normativity, see James W Harris, “Kelsen’s ‘Pallid Normativity’” (1996) 9 Ratio Juris 94. 319. Joseph Raz, The Authority of Law (Oxford University Press, 1979) at 154 [Raz, Authority of Law]. 320. Ibid at 134-35, 153-59. For an explanation, see also Julie Dickson, “Interpreting Normativity” [Dickson, “Interpreting Normativity”] in T Endicott, Joshua Getzler & Edwin Peel, eds, Properties of Law: Essays in Honour of Jim Harries (Oxford University Press, 2006) 22. See also Raz, Authority of Law, supra note 319. See also Gerald Postema, “The Normativity of Law” in Ruth Gavison, ed, Issues of Contemporary Legal Philosophy: The Influence of H.L.A. Hart (Clarendon Press, 1987) 81 at 81, 83ff. 321. The example might be stretched, as in religious law more than anywhere, it matters who the persons of adviser and advisees are. On Islamic law, see Lawrence Rosen, The Justice of Islam: Comparative Perspectives on Islamic Law and Society (Oxford University Press, 2000). 322. Raz, Authority of Law, supra note 319 at 140, 156, citing Kelsen, supra note 37. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 491 system.323 Adopting this point of view, legal scholars and practitioners are able to make normative statements that do not convey their approval of the norms articulated therein.324 For Raz, these statements are properly normative: They are not external descriptions of people’s beliefs, attitudes, and actions, but statements about what is to be done from that legal point of view and asserting what legal rights and duties people have.325 In short, they are statements of (and not just about) law. However, detached statements do not carry the full normative force of those uttered from the internal perspective—what Raz calls “full-blooded” statements—because the speaker is not equally committed to the view that they express.326 They are normative in Raz’s particular sense, “statements about what there is reason to do from the point of view of a normative system to which they are not committed.”327 Even though the argument is not as straightforward as Raz would have it,328 it is not hard to see what he is driving at: One need not subscribe to a normative position in order to state that someone ought to follow it as valid law. Before assessing this argument, it is necessary to explain the role that detached normative statements are meant to fulfill in Raz’s general theory of law. Raz’s main positivist thesis is that law is “source-based,” meaning that the existence and validity of legal rules can be “identified by reference to social facts alone, without recourse to any evaluative argument.”329 For Raz, the existence or validity of legal rules depends on their source, which is established in conformity with tests laid down by some other rules of the system rather than by “argument Ibid at 141. Ibid at 156. Ibid at 153. Ibid at 153, 156. Raz states that they are “parasitic on full-blooded normative statements” (ibid at 159). 327. See Dickson, “Interpreting Normativity,” supra note 320. 328. By speaking of statements that are simultaneously detached (non-evaluative) and normative (i.e., evaluative though not in a “full-blooded” sense), the impression is hard to avoid that Raz wants to have the cake and eat it too. See Roger A Shiner, Norm and Nature: The Movements of Legal Thought (Oxford University Press, 1992). Shiner argues that even though Raz adopts the language of J.L. Austin to speak about the normative “force” of the statement, he assumes the possibility of a declarative statement which stands by itself and to which “force” is added later on, which Austin explicitly rejected. In Austin, “What makes a use of language ‘normative’ is that someone uses a given locution to perform the illocutionary act of making a normative judgment” (ibid at 143). Therefore, if detached legal statements do not have full normative force, they are not normative (ibid at 145). 329. Joseph Raz, “Authority, Law and Morality” (1985) 68 Monist 295 at 296. 323. 324. 325. 326. 492 (2022) 59 OSGOODE HALL LAW JOURNAL concerning [their] value and justification.”330 Given that law is a “social fact,” “doubts and discussions about the validity of laws revolve on factual questions... on issues susceptible of objective determination to which one’s moral or political views are essentially irrelevant.”331 This source-based conception of law is “vital,” Raz has claimed, for “it alone can guarantee that the content of law can be determined in an objective and value-neutral way.”332 Such is the purpose of detached normative statements: to ensure that the determination of what the law is—what obligations, duties, rights, et cetera people have—can be done in a matter-of-fact, objective way, without recourse to evaluation. From a Bakhtinian perspective, Raz’s attempt is unsuccessful on three different grounds: first, it is not possible to state what the law is without engaging with deeply evaluative considerations; second, there is no such thing as a system’s “legal point of view”; and third, what Raz calls “detached” is actually a form of dialogical discourse that cannot be apprehended in the declarative and propositional sentences of positivism. All these lead to the conclusion that legal positivism fails in its basic goal of providing a factual, non-evaluative determination of the law. I will pursue this argument with the help of Raz’s own simplified example of the Catholic who offers legal advice to an orthodox Jewish friend, which is shown to have no basis in the actual experience and responsibility of ascertaining the law. I conclude with an example of a detached normative statement drawn from a legal case, which demonstrates that this form of statement typically lacks awareness of the loaded normative charge it carries—which has serious detrimental effects for the kind of neutrality that legal positivism seeks to defend. My general point is that we need a different language—in fact, an entirely dialogical conception of language—to do justice to complex legal utterances such as these. In Raz’s theory, detached normative statements are concerned with identifying valid rules of the legal system—not so much with the propositions and imperatives (or the “deontic content”) they contain.333 They are concerned with the empty shell of the rule, the content of which the speaker does not appear to endorse merely by stating it. But to state that a rule is valid is not to say that 330. Raz, Authority of Law, supra note 319 at 150-51. Raz believes that the validity of the rule is the same as its existence (the rule exists if it is valid and the rule is valid if it exists). Further, a rule is valid if it is binding, that is, if it has the normative consequences it purports to have. See ibid at 149-50, 153. 331. Ibid at 152. 332. Ibid. 333. Ibid. For Raz, rules are to be distinguished from the propositions or imperatives (the “deontic content”) contained therein. In this view, rules “are not statements nor prescriptions….They are things the content of which is described by some normative statements” (ibid at 148). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 493 it should apply—the rule can be perfectly valid and yet inapplicable for a series of circumstances: that it is not expedient to follow, that it may bring unwanted consequences, that there are countervailing norms and principles that oppose it, etc. (We have seen this when we tested “No vehicles in the park” as an empty-shell rule.) Even in a simplified model of rules that positivism proposes, then, stating the law is not about identifying valid rules, but about assessing their tenor, scope, applicability, possible exceptions, counterclaims, and so on. This means that one cannot ascertain what legal rights and duties people have without engaging in myriad evaluations. Ascertaining the law is not a matter of fact but of judgment,334 and it requires utterances, not sentences. In Raz’s example, in order for the Catholic friend to ascertain the obligations of a practicing Jew—to state what obligations the friend has in a particular situation—it is not enough to point to a valid rule. Our Jewish friend does not want a restatement or tautology but legal advice as to what to do in a particular situation, which requires assessing possible exceptions to the rule, differences of opinion within diverse Rabbinic schools, complex casuistic narratives, and so forth. As a result, the Catholic friend cannot provide legal advice without getting bogged down in the complexities of the case and the legal tradition—how else can the Catholic friend demonstrate “expertise” on Jewish law? This raises the prior, non-obvious question of expertise that Raz’s example simply assumes. If I were to speak as a Catholic trying to offer legal advice to my “not-so-knowledgeable” Orthodox friend, I would want to make sure above all that I do not impose my own different views—replacing my friend’s values with mine. I would therefore be careful that my different presuppositions and background assumptions do not stand in the way of the advice, either by clouding my judgment or by distorting the legal tradition of my friend, being appropriately respectful of orthodox Judaism as a way of life. I would also want to make sure that I understand what being an orthodox Jew means for my friend specifically and the role Judaism plays in my friend’s life: the significance attributed to this or that practice, this or that ritual, this or that scripture, this or that communal act, et cetera. In other words, my legal advice would be mediated by my understanding of the Jewish legal tradition, my specific friend, and our mutual friendship. To honour these, I would avoid presuming too much about my own expertise; I would offer advice with humility and entertain the possibility of error and personal shortcomings; I would also be sensitive to the circumstances of my friend. All of these would affect my exact words of advice. 334. Timothy Endicott, “Law and Language” (15 April 2016), online: Stanford Encyclopedia of Philosophy <plato.stanford.edu/entries/law-language>. 494 (2022) 59 OSGOODE HALL LAW JOURNAL For example, I would probably not say “this is what you ought to do,” but rather, “here is how I understand your obligations and how I think you can follow them too.” In asserting this, I do not just state a legal rule but establish a relationship between myself, my friend, and my friend’s law. In any event, I would assume full responsibility for my advice—and not hide behind a “detached” form of impersonal advice. Does this mean that I would need to convert to Judaism to be able to advise my Jewish friend? Surely not. But a certain sensibility, open receptivity, and attunement are vital.335 For instance, without an appreciation of the sacred and unbreakable bond of the Covenant between the Israelite people and God, I would hardly be in the position to offer any sound legal advice. This is to say that the kind of expertise required is not simply a matter of cognition but also of disposition, perception, and sensibility. Here we may recall Marmor’s example about recognizing the importance of Catholicism in Baroque architecture without having to be appreciative of either Catholicism or Baroque art. It is true that I do not need to be a practicing Catholic to be able to appreciate Baroque art, but I would certainly need to develop some appreciation for the sacred and metaphysical aspects of Catholicism; its symbology; its conceptions of time and space, heaven and hell, light and darkness, matter and body; and for the carnal and mystical imagery of suffering, torment, and ecstasy in order fully to appreciate Baroque art. Therefore, one cannot “grasp” these values without having some evaluative orientation, because grasping a value is not simply an intellectual or cognitive process. Similarly, to determine what the law is in a given situation requires a sense of orientation, perceptiveness, acumen, and practical experience. Legal expertise is not just a book type of learning. This takes us to the second main objection regarding Raz’s so-called legal point of view—the position of the hypothetical “legal man” who accepts all and only the laws of their country. According to Raz, legal practitioners and law professors speak from this perspective, from “the system’s own point of view.”336 The objection is straightforward: there is no such thing as “the system’s own point of view.” I do not mean to say that “the legal man” is a non-existent 335. On the importance of receptivity, see Jennifer Nedelsky, “Receptivity and Judgment” (2011) 4 Ethics & Global Politics 231. See also Rita Felski, The Limits of Critique (Chicago University Press, 2015); Nikolas Kompridis, Critique and Disclosure: Critical Theory Between Past and Future (MIT Press, 2006) at 199-210. On attunement as a form of justice, see Richard Dawson, Justice as Attunement: Transforming Constitutions in Law, Literature, Economics and the Rest of Life (Routledge, 2014). See also Becker, supra note 2. 336. Dickson, “Interpreting Normativity,” supra note 320 at 39. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 495 person—after all Raz himself admits that this is a hypothetical construct337—but this explanatory device is premised upon a reified notion of the legal system that can never be realized in actual life and practice. The so-called legal point of view assumes an external body of law that can be accessed from a neutral perspective—the supposedly “transparent interpreters” of texts that “speak on their own”338—but the existence of neither a disembodied law nor a neutral perspective can be taken for granted. Since the reception of the Corpus Juris Civilis to the heyday of legal positivism, it has been common to view law as a reified object—having its existence in written or legislated form. This notion may seem well-suited to the “religions of the book” in the Jewish example. Contrary to appearances, however, there is no body of Jewish law that I could consult independently from the practices, beliefs, narratives, and personal commitments of practicing Jews.339 Even when a legal tradition seems quite literally to be contained in a book, then, in order to apprehend its normative valence, the book would have to “come to life.”340 As living language, law cannot be disembodied or depopulated. In this regard, Raz argues that detached normative statements are not about people’s beliefs and hence obtain irrespective of anyone actually holding the beliefs.341 But if there is no one actually to hold them—if it is irrelevant in fact whether anyone ever did or will—then how are “experts” to know if they are accurately representing “the legal point of view,” especially when they are not personally committed to it? Legal statements without actual people questioning their application, discussing their scope, qualifying their reach, arguing about their meaning, and disputing 337. Ibid at 38. 338. See Section V, above. 339. See Barry Wimpfheimer, “Codes” in Julen Etxabe, ed, A Cultural History of Law in Antiquity, vol 1 (Bloomsbury, 2019) 59. 340. On the creativity of Biblical interpretation, see Bernard M Levinson, “The Right Corale”: Studies in Biblical Law and Interpretation (Mohr Siebeck, 2008). 341. See Raz, Authority of Law, supra note 319. Raz claims that “there may be no one who has such a belief ” (ibid at 157) [emphasis added]. It is one thing to say that the statement does not represent beliefs of any single individual; another quite different thing is to say there is no need for anyone whatsoever to hold any such belief. Either I am missing something here, or the argument is seriously inadequate. 496 (2022) 59 OSGOODE HALL LAW JOURNAL their normative content are as hollow as a language without speakers: They cease to exist as living discourse to become the dead sentences of grammarians.342 Finally, detachment is understood by Raz as entailing no personal involvement whatsoever.343 However, as the early parts of this article have shown, statements of law are utterances rather than empty shell grammatical units (or sentences). As such, a speaker cannot help but to intone an utterance, choose an appropriate register for it, a suitable form of address, a fitting style, a genre—and await a response. In doing so, a speaker puts forth a world of value that is offered for assent or criticism; adopts certain terms of commendation or reproach; assumes certain aims and ideals to be shared or disputed; invites certain ways of arguing, of drawing analogies, and of giving examples. All of these betray profoundly evaluative orientations. For example, even if an anarchist law professor would be able to describe positive law as valid as Raz and Kelsen are fond of stating it is unlikely that the description would not be tainted by irony, disdain, critical outlook, parody, or contempt—or even a latent invitation to resist. Aren’t we missing something essential at a descriptive level when we fail to distinguish the legal utterances of an anarchist from those of a legal apologist? *** Some readers may still find Raz’s notion of detached statements appealing because we do not always seem to endorse the normative views that we are able to describe. The error in this view, however, is to believe that detached statements entail no personal involvement of the speaker. At the very least, a statement purporting to declare the law in a given situation commits us to the accuracy of the description we are making and, additionally, to the assumption that the other person too is bound by it—which ultimately depends on how they answer. To state something in a “detached” way is, therefore, to open up a dialogical space between the statement and its normative consequence, which the propositional conception of language fails to register. 342. Ibid. Raz seems to be conscious of this. He writes that “there is normally no point in making statements from a point of view unless in relation to a society in which people are often ready to make the full-blooded statements. If there is nobody whose point of view it is, why should we be interested in it?” (ibid at 159). But the issue is not one of interest, but of ability: How do we know we have correctly identified their obligations if there is actually no one “whose point of view it is”? 343. For a critique, see Luís Duarte D’Almeida, “Legal Statements and Normative Language” (2011) 30 Law & Phil 167. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 497 The point can be clarified with an example taken from Neil MacCormick, slightly modified here.344 Suppose that I want to describe the normative universe of Nazism and one of the things that I state is that, for the Nazis, racial discrimination is obligatory. MacCormick suggested that “my statement can count as detached only if it is clear that I convey no opinion as to the correctness of the view, and that my view can differ.”345 That is, I “refrain…from having any view of [my] own.”346 This sounds puzzling: how can I express “no opinion” and, at the same time, make it “clear” that “my view can differ” from it? It becomes apparent that in order to express my “detachment” (i.e., separation) from it, I must be able to create a wedge between my current statement (declaring that racial discrimination is obligatory) and the views of the Nazis I try faithfully to represent, lest my position be misunderstood and confused with theirs. Otherwise, how can I clarify that my viewpoint is not necessarily the same as, and can differ from, that of the Nazis? Arguably, a “detached statement” is not to utter a “no-opinion” but to utter an opinion that, however subtly, differs substantially from the views I now try to represent. In short, to utter a detached point of view does not mean to “refrain…from having any view of one’s own”;347 it means establishing a “double-voiced” position, putting a buffer between my opinion and the one I try to represent to the best of my abilities. Thus interpreted, the statements of the critic unpacking Nazi ideology, the anarchist law professor, and the Catholic friend are all forms of double-voiced discourse. Rather than putatively accessing a neutral “legal point of view,” these utterances are internally dialogized, that is, mediated by the speaker’s appreciation of someone else’s obligations, duties, rights, and so on.348 The ability to perceive 344. Neil MacCormick, “Comment” in Ruth Gavison, ed, Issues of Contemporary Legal Philosophy: The Influence of H.L.A. Hart (Clarendon Press, 1987) 105 at 109. 345. Ibid [emphasis in original]. 346. Ibid [emphasis in original]. 347. Ibid [emphasis in original]. 348. Neil MacCormick calls this the “hermeneutic point of view.” H.L.A. Hart (Stanford University Press, 1981). Echoing Raz, MacCormick thinks that “normative statement may be made either from the internal point of view or from the hermeneutic point of view” (ibid at 39 [emphasis in original]), the former requiring both volition (will) and cognition (understanding), the latter only understanding (not volition). See Neil MacCormick, “Appendix” in Legal Reasoning and Legal Theory (Oxford University Press, 1978). See especially, ibid at 288ff [MacCormick, “Appendix”]. For MacCormick, the hermeneutic perspective comes determined by “the understanding, not the will of the speaker” (ibid at 291 [emphasis in original]). However, this distinction is not without its problems: it presupposes that the mind can be divided up into divided elements, and in particular that cognitive and volitional elements may be systematically distinguished. It presupposes also that 498 (2022) 59 OSGOODE HALL LAW JOURNAL and represent the normative view of others demands what Bakhtin calls “un-indifferent” or “participative thinking” (not to be confused with empathetic identification), which is not merely a cognitive appraisal.349 Because I cannot understand an utterance without having an evaluative orientation towards it, any statement of law necessarily incorporates my assessment of the normative valence of the situation. I want to illustrate this last point with a real example. The case revolves around the process of extradition of General Augusto Pinochet when UK authorities decided to arrest him while he was on a short trip to London in 1998. The double arrest warrant issued by Spanish authorities on the charges of genocide and torture (among others) stated that, after the 1973 coup d’état to overthrow the government of Salvador Allende, Pinochet had ordered the elimination, disappearance, and kidnapping of thousands of persons who were also subject to systematic torture.350 After the arrest, a writ of habeas corpus was filed on behalf of Pinochet before a three-judge panel of the UK High Court. The claim was that, as a former head of state, Pinochet was immune from prosecution, and therefore the arrest was invalid. Acting on behalf of Spain, the Crown prosecutor argued that some crimes were so deeply repugnant to any notion of morality as to constitute crimes against humanity, and there could be no immunity in respect of them. The Divisional Court had earlier sided with Pinochet’s defence and decided to quash the arrest warrants pending an appeal, arguing that Pinochet had immunity for crimes committed during his tenure as Chile’s head of state.351 Rather than focus on this judgment and on the tortuous legal battle that ensued before the House of Lords, I want to highlight a short statement made by one of the members of the Divisional Court, which best exemplifies the problem of so-called detached normative statements. For Justice Collins, the whole case depended upon establishing that the applicant (Pinochet) was acting as head of government, in which capacity he enjoyed immunity. The prosecutor submitted that “it could never be in the exercise of such functions to commit crimes as understanding the nature of rules and of rule following behavior is in particular a state of mind without a volitional dimension (Shiner, supra note 328 at 146). 349. See Bakhtin, Act, supra note 142. As Bakhtin writes, “An event can be described only participatively” (ibid at 32). 350. The British authorities received two different warrants because the first one was defective. On the legal vicissitudes surrounding the case, see Naomi Roht-Arriaza, The Pinochet Effect: Transnational Justice in the Age of Human Rights (University of Pennsylvania Press, 2005). 351. United Kingdom High Court of Justice, Queen’s Bench Division (Divisional Court), ‘In the Matter of an Application for a Writ of Habeas Corpus ad Subjicendum. Re: Augusto Pinochet Duarte’” (1999) 38 ILM 68 at para 79 [Habeas Corpus]. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 499 serious as those allegedly committed by the applicant.” Justice Collins responded in the following manner: Unfortunately, history shows that it has indeed on occasions been state policy to exterminate or to oppress particular groups. One does not have to look very far in history to see examples of that sort of thing having happened. There is in my judgment no justification for reading any limitation on the nature of the crimes committed into the immunity which exists.352 Justice Collins also stated that, due to the seriousness of the matter and the terrible crimes attributed to the defendant, there might be a “great temptation” to send Pinochet to Spain to be tried, but insisted that “one cannot twist the law to meet the apparent merits of any individual case.”353 My aim here is not to discuss the legal complications of the case or whether Justice Collins was justified or not in reaching that decision.354 Nor is it to rehearse the well-known debate about the obligation to (dis)obey an unjust law.355 More pointedly, I want to take Justice Collins’s response as precisely the type of (supposedly) detached normative statement of which Raz speaks—that is, a statement of law in which the speaker claims no personal commitment to the law being expressed. That this particular assertion is made by a judge and not by a legal scholar makes no difference as to the kind of statement it purports to be, namely, a declaration about what the law is on the matter.356 The judge purportedly did not speak from his personal perspective but from “the point of view of the legal system.” According to Raz, the judge need not be personally committed to the view that he stated, nor approve of the law in order to state it as obligatory. Finally, the judge’s statement appears to convey the paradigmatic positivist position: The merit or demerit of the law is irrelevant to determining the 352. Ibid. 353. Ibid. 354. Some of these arguments would involve discussing the 1948 Genocide Convention and its transposition to the UK in 1969 and the 1984 Convention Against Torture (UK 1988). The discussion would also need to interpret the key historical precedents of Nüremberg and the International Criminal Tribunals for the Former Yugoslavia and Rwanda. Furthermore, it would need to acknowledge the rule of immunity as not absolute, and that the final decision on immunity could have been left to the Spanish authorities. 355. For an interesting case study in the context of Apartheid in South Africa, see David Dyzenhaus, Judging the Judges, Judging Ourselves: Truth, Reconciliation and The Apartheid Legal Order (Hart, 1998). On disobedience more generally, see Candice Delmas, A Duty to Resist: When Disobedience Should be Uncivil (Oxford University Press, 2018). 356. There is obviously an added component on the ability of the judge to make his statement effective, but the question facing both the judge and the scholar would not substantially differ. 500 (2022) 59 OSGOODE HALL LAW JOURNAL content of the law. In short, this seems a perfect example of detached normative statements that convey no personal viewpoints. In what ways does such a description fail to capture Justice Collins’s remarkable utterance? I suggest, quite simply, that it fails to capture the heavy normative work it is performing to validate the law as interpreted. The judge’s task is not merely to declare that the rule of immunity exists and is applicable but to demonstrate that the reasons for it are weightier than the reasons opposing it. This issue cannot be settled simply by pointing to the existence of a rule, because it must be ascertained whether, and how far, immunity is to apply in the contested circumstances. The judge had to assess whether the behaviour that Pinochet was accused of having committed falls within the scope of the rule. Further, he had to decide whether immunity should prevail in the face of a constellation of countervailing norms, historical narratives, and geopolitical developments to the contrary. To buttress the argument for immunity by pointing to recent historical examples of state policies designed to exterminate or to oppress certain groups of people does little to advance the proposition that this kind of behaviour should be legally protected. Rather the opposite: It serves as an unfortunate reminder of why immunity was being challenged in the first place. In offering these examples, the judge invited us to view acts of mass torture and forced disappearance as instances of the rule, i.e., as part of the acceptable range of actions performed by heads of state deserving immunity. This seems an unusual claim for a rule that is said to derive simply from the comity of nations.357 It was also curious that the judge gave these historical examples normative credence, without an attempt to explain why they should continue to bind our current understanding of the norm.358 Positivists may respond that the judge was expressing no personal view on the matter, but only that of “the law.” But because no interpretation is immutably written in the rule itself, the judge chose a version of the rule that obscured his choice by presenting it as the opposite of a choice. Furthermore, he invited his audience to accept particularly egregious instances of the application of the rule, such that any attempt to draw limitations on immunity due to the nature of the crimes committed is dismissed as having “no justification.” Finally, his statement 357. The Lord Chief Justice (Lord Bingham) explained the rationale for immunity to be “a rule of international comity restraining one sovereign state from sitting in judgment on the sovereign behavior of another.” Habeas Corpus, supra note 351 at para 63. 358. It should be stressed that my objection has nothing to do with the Humean capital sin of deriving an “ought” from an “is.” In a Bakhtinian understanding of language, these categorical distinctions do not hold. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 501 invites us to admire his resolve not to fall into the temptation of “twist[ing] the law to meet the apparent merits of the case,” as if discussing the particular merits of a case would be foreign to the task of adjudication. It would be grossly unfair to suggest that Justice Collins wished to condone the kind of behaviour Pinochet was accused of having committed. In this regard, it may be right to say that his declaration does not necessarily assess the merit or demerit of the charges. Still, his statement is the opposite of detached—or it is so in the very different sense that it disavows the normative consequences it necessarily entails.359 Instead, it shows commitment to a particularly monological version of legal positivism, which risks turning into a practical defence of, and apology for, legalism. VIII. THE GRAMMAR OF LAW AND THE GAME OF CHESS Throughout this article, I have argued that legal positivism entails, and builds upon, an “abstract objectivist” conception of language. I want now to explicitly highlight the through-line from linguistics to legal positivism by examining the use of a popular metaphor within both fields of study, namely the trope of comparing their respective objects of study to the game of chess.360 Hart famously referred to chess at various points in The Concept of Law to illustrate his key distinction between the internal and the external aspects of rules.361 It is perhaps less well-known that chess was also a favourite for Saussure: “[O]f all comparisons that might be imagined, the most fruitful is the one that might be drawn between the functioning of language and a game of chess…[for a] game of chess is like an artificial realization of what language offers in natural form.” 362 The chess metaphor can thus be seen as the point of intersection for the theory of language and the theory of law of which Saussure and Hart were such influential proponents. By looking closely at the jurisprudential work that this metaphor 359. This sense of detachment as disavowal brings it closer to the way “detachment” is understood in psychoanalysis as denial, deflection, and displacement (I thank Peter Goodrich for this remark). 360. For a more recent example, see Andrei Marmor, “How Law Is Like Chess” [Marmor, “Chess”] in Law in the Age of Pluralism (Oxford University Press, 2007) 153. See also Arie-Jan Kwak, “What’s in a Game? Legal Positivism as ‘Still Descriptive and Morally Neutral’” (2017) 2017 L & Method 1. 361. Hart, The Concept of Law, supra note 173 at 56. And before him, the leading Danish legal philosopher, see Alf Ross, On Law and Justice, ed by Jakob VH Holtermann, translated by Uta Bindreiter (Oxford University Press, 2019) at 20-24. If one can doubt that Hart had read Saussure, Carnap, and Wittgenstein, there is no doubt that Hart had read Ross. 362. Saussure, supra note 28 at 88. 502 (2022) 59 OSGOODE HALL LAW JOURNAL does, I aim to show that it is tilted towards the construction of a particular object of study to which legal positivists are committed. It is generally believed that Hart’s understanding of language was influenced by ordinary language philosophers, who often used aspects of common life, including games, to illustrate their arguments.363 Even earlier, Rudolf Carnap, a prominent figure in the logical positivism of the Vienna Circle during the 1920s and 30s, made a key distinction between two types of language rules that may be familiar to readers of Hart, also using chess to draw an analogy. According to Carnap, “the rules of formation determine the position of the chessmen (especially the initial position in the game), and the rules of transformation determine the moves which are permitted—that is to say, the permissible transformations of one position into another.”364 The parallels between Carnap’s rules of formation/transformation and Hart’s secondary rules reveal an alternative genealogy for the chess metaphor that links legal positivism even more closely to the emergence of modern linguistics. For Saussure and the structural linguists influenced by him, it was important that the study of language be scientific. This presupposed the ability to distinguish what is intrinsic (or proper) to language from what is extrinsic (or foreign) to it. To explain the difference, Saussure uses the game of chess to offer a defining analogy: As we have already mentioned, playing the game with chess pieces made of different materials would make no difference to the game, whereas a change in their number would dramatically alter the game. While the former is an extrinsic factor, the latter intrinsically affects the system and hence falls within the sphere of linguistic study. This is because, for Saussure, language (langue) is an interdependent whole in which each word, like pieces in chess, depends on its relative position on the chessboard, and above all on the unchanging rules of the system that preexist any particular occurrence of the game and continues after it has ended.365 Language, like chess, pre-exists any particular player and continues intact after each play. The convergence between structural linguists and legal positivists is enlightening and reveals four underlying assumptions driving legal positivism as a theory of law. First, the game of chess is not used to emphasize the ludic or 363. Ludwig Wittgenstein, Philosophical Investigations, ed by PMS Hacker & Joachim Schulte, 4th ed, translated by GEM Anscombe (Wiley-Blackwell, 2009), ss 197, 205 (“But isn’t chess defined by its rules?” wondered Wittgenstein, suggesting that “chess is the game it is by virtue of its rules”). 364. Rudolf Carnap, Logical Syntax of Language, translated by Amethe Smeaton (Routledge, 2001) at 5. 365. See Saussure, supra note 28 at 88. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 503 playful aspects of human interaction.366 Rather, chess provides a simplified model of a highly structured activity with well-defined and clearly articulated rules. According to Hart, legal rules define not only the appropriate moves but the participants’ internal perspective when moving a piece: Those who accept the rule do not merely have the habit of moving the piece in a certain manner, but regard this movement as “the right move.”367 For Hart, this difference must translate into the perspective of the jurisprude as well, who is to distinguish action that is genuinely an observance of a rule from one that merely happens to coincide with it.368 As a result, anyone wanting to describe what participants are doing must realize that participants move pieces not out of habit but in accordance with certain rules.369 Secondly, the metaphor of chess is employed to suggest not only that the game is rule-governed, but that certain rules are so fundamental that, without them—or if they were suddenly to be replaced or altered, as in Saussure’s example of a sudden increase in the number of chess pieces—the game would cease to be the same. In this vein, Hart rhetorically asks: “Is it still ‘chess’ if the game is played without a queen?”370 A few years later, John Searle employed the chess metaphor to exemplify his key distinction between constitutive and regulative rules, which has become a hallmark for positivists.371 According to Searle: [R]egulative rules regulate antecedently or independently existing forms of behavior; for example, many rules of etiquette regulate inter-personal relationships which exist independently of the rules. But constitutive rules do not merely regulate, they create or define new forms of behavior. The rules of…chess, for example, do not merely regulate playing…chess, but as it were they create the very possibility of playing such games. The activit[y] of playing…chess [is] constituted by acting in accordance with (at least a large subset of ) the appropriate rules.372 Building on Hart and Searle, Andrei Marmor has made the connection between chess and law explicit, suggesting that the basic rules of the legal system, the so-called rules of recognition, “are very much like the rules of chess: they 366. On the importance of this ludic element, see Johan Huizinga, Homo Ludens: A Study of The Play-Element in Culture (Routledge, 1949) at 76-88. On the lack of “play” in structural thinking, see Derrida, Writing, supra note 236 at 278-93. 367. Hart, The Concept of Law, supra note 173 at 288, n 54. 368. Ibid at 140. 369. Ibid at 89. 370. Ibid at 4. 371. See Fernando Atria, On Law and Legal Reasoning (Hart, 2002). 372. Searle, supra note 175 at 33-34 [citations omitted]. 504 (2022) 59 OSGOODE HALL LAW JOURNAL constitute ways of creating law and recognizing it as such.”373 Marmor admits that chess is not only a matter of following rules—people play for many and varied reasons—but the rules determine the “concept” of the practice.374 It is in this sense that these rules are constitutive: They define the game for what it is. In other words, the game could not be identified but for the existence of these rules. The analogy is thus clear: Without the social conventions that constitute ways of making law and recognizing it as such, it would be “difficult to imagine what kind of concept of law we could possibly have.”375 In this view, just as the rules of chess constitute the point of playing the game and some of the specific values associated with it, the rules of recognition are the “social conventions that determine what counts as law in a given community.”376 The third feature of the chess metaphor as used by positivists is that it sets up an external object of observation that remains unaffected by individual acts or actors. Just like Saussure invites us to keep our eyes on what remains constant (langue) rather than on what is contingent (parole), legal positivists emphasize the game that pre-exists and continues after each play. Even though there might be different ways of playing chess (and law), variations in style or slight alterations of non-fundamental rules do not change the nature of the game. In addition, the analogy helps to establish a fairly sharp demarcation between players, spectators, and other nonparticipants, as well as a certain attitude of detachment from real-life concerns (i.e., the knowledge of playing “just a game”). Fourth and finally, the analogy depicts law as a bounded whole. For positivists, legal sentences or propositions are not isolated entities per se but the most elementary units of a larger system, such that it is possible to identify whether a particular legal proposition is correct or incorrect, valid or invalid.377 As explained by Norberto Bobbio, spiritual founder of the Italian analytical school, “At the basis of every legal system there is a fundamental rule according 373. Marmor, “Chess,” supra note 360 at 166. 374. Ibid at 170. 375. Ibid. Marmor suggests that the rules of particular games are to be read against the background of deeper conventions about playing competitive games in general, thus making a distinction between “surface” and “deep” conventions in law too. For a theory of law developed along the lines of deep and surface layers, see Kaarlo Tuori, Critical Legal Positivism (Ashgate, 2002). 376. Marmor, “Chess,” supra note 360 at 172. 377. Manuel Atienza and Juan Ruiz Manero write that “[l]egal sentences are…the most elementary units of law; but those pieces acquire full meaning only when their contribution to the shape and functioning of the law is well understood” (supra note 8 at xi). In their “taxonomy of the types of [legal] sentences,” law is “seen as langue…and not as parole (ibid [emphasis in original]). ETXABE, THE DIALOGICAL LANGUAGE OF LAW 505 to which the group of normative propositions making up the system constitutes a closed whole.”378 This fundamental rule functions like the unifying grammar of law, “transform[ing] a set of propositions into a system.”379 Knowing this “grammar” is to know how these various rules operate within a system that participants must assimilate as a given.380 The jurist integrates a particular proposition into the normative system by using the rules of transformation admissible for that system, or alternatively, by removing those propositions which are not similarly deducible.381 Consequently, “legal analysis is conducted within the narrowly circumscribed limits of a particular language,”382 or grammar.383 The centrality of the chess metaphor in conceiving a theory of law has far-reaching consequences. For legal positivists, the analogies derived from chess enable them to describe law as a rule-governed activity and, in fact, as constituted by rules. Law can thus be studied independently from the behaviour of legal actors. The proper object of a general theory of law is to describe this system. As Peter Goodrich notes, the legal order thus conceived can be matched “detail for detail” with the Saussurian science of linguistics.384 But what if law is nothing like chess and the metaphor is leading legal positivists astray? First, when we say that the game of chess consists of following rules, what are we actually describing? Suppose that someone who had never heard of the game approached us with the question “what is chess?” and we proceeded immediately to explain the different movements of the pieces. What about the game do the rules actually capture? Certainly, rules do not explain 378. Norberto Bobbio, “The Science of Law and the Analysis of Language” [Bobbio, “The Science of Law”] in Pintore & Jori, supra note 8, 21 at 43 [emphasis in original]. 379. Ibid. 380. In an insightful and quite congenial article, Ingo Venzke associates law’s grammar with Chomsky’s structural linguistics (and his theory of generative grammar) more than with Saussure. See “Is Interpretation in International Law a Game?” in Andrea Bianchi, Daniel Peat & Matthew Windsor, eds, Interpretation in International Law (Oxford University Press, 2015) 352. While there are important differences between Saussure and Chomsky, I simply wish to point out their structural similarities. 381. Bobbio, “The Science of Law,” supra note 378 at 43. 382. Ibid. 383. Ibid at 41. Interestingly, the emphasis on law’s grammar is not exclusive to legal positivism. From a critical perspective in international law, expressing an underlying linguistic theory closer to Chomsky, Martti Koskenniemi argues that “how international lawyers argue... can be articulated in a limited number of rules that constitute the ‘grammar’–the system of production of good legal arguments.” See From Apology to Utopia: The Structure of International Argument (Cambridge University Press, 2006) at 568. 384. Goodrich, “Law and Language,” supra note 20 at 181. See also Goodrich, Legal Discourse, supra note 6 at 35ff. 506 (2022) 59 OSGOODE HALL LAW JOURNAL how the players experience the game, why fans enjoy it, or why anyone would be interested in the game in the first place. This is to say that a focus on the rules is neither the most interesting nor the most enlightening description of the game. At most, we would be adopting the perspective of a referee trying to explain the rules from a manual to a novice—although it is not clear why we should privilege such a perspective. To be able to parrot the rules does not help us to understand the players, the role that the game plays in the wider world, the specific values attached to playing the game, or why people are (or are not) drawn to it. The first criticism is simple: Focusing on the rules of chess may distort what playing it is all about. A positivist may want to respond that law is constituted by the rules in the sense that the game of chess cannot be understood but for the existence of these rules. But here’s a second problem. On its own terms, the distinction between regulative and constitutive rules is less apparent than what Searle and Marmor care to admit. On its face, I fail to see what makes the rules of etiquette regulative whereas those of chess are said to be constitutive. It would appear that etiquette makes no sense but for the existence of some rules of etiquette: How to pick up a fork or what dress to wear to a formal party are not things that make much “antecedent” sense, except as part of a complex web of norms and social conventions that precisely define etiquette. Even if the distinction is taken at face value, there are still two possible senses of the term “constitutive” that make law unlike chess. On the one hand, the rules of chess (like the rules of competitive games in general) are not meant to stand in the way of the game. They may define the game for what it is, but they are meant to stay out of sight for the sake of the game because it is the game and not the rules that matter to participants and spectators alike. We may say that the rules have a “vocation to recede”: Only when there is a problem is a rule invoked, and even then, the judge or arbiter often responds tautologically, merely indicating the existence of the rule, without offering further elaboration or development. No precedent or authority needs to be cited other than the rule itself; no attempt needs to be made to justify it. On the other hand, constitutive norms as known in law are quite different. They are the ones we live by and that help to construct the normative universe in which we live.385 Examples abound of rules that fulfill a constitutive role in their respective legal universes: rules guaranteeing secularism and the democratic character of the state; respect for human dignity; collective bargaining; trial by jury; the obligation to consult; and self-determination. These are not mere “general principles of law,” but rather the backbone of social and legal identities, 385. Most expressively, see Cover, supra note 58. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 507 in ways that resemble precepts in some religions (such as iconoclasm, observance of Sabbath, prohibitions on certain foods, et cetera). Such norms do not recede once the activity starts but continue to have a sustained presence in the lives of the individuals and communities that decide to make them their own. We may say that constitutive norms of this second kind have a “vocation to persist.” In other words, these norms do not give way to the activities themselves but remain visible and central for the duration. People talk about the norms and make them relevant in their ordinary lives. People discuss their significance, their origin and justifications, their merits and demerits, their implications, all the while dynamically interacting with them in ways that can help to make sense of their normative experience. People invoke such constitutive norms not only when there is a problem or a case of conflict. In fact, when a case is brought to an adjudicator, the disputants usually expect more than a tautological invocation of a rule; they expect a legal judgment that may cite precedents and consider alternative justifications of the norm. My argument is that legal positivists understand constitutive rules in the first sense, but not in the second. Paradoxically then, even though legal positivism emphasizes rules over all other possible aspects of law, when it comes to this kind of value-laden, normatively charged constitutive norms, legal positivists do not fully grasp the role that such norms play in the lives of individuals and groups. If, as Robert Cover famously argues, “Law is a resource in signification that enables us to submit, rejoice, struggle, pervert, mock, humiliate or dignify,”386 then comparing legal norms with the rules of chess misses this rich significance.387 A third distorting effect of the chess metaphor is that gives the illusion of completion and mastery. Conceiving of law as chess gives the impression that law can be reduced to a limited number of rules which determine all possible combinations and permutations. By contrast, a Bakhtinian understanding of legal language is inexhaustible.388 We may develop more or less familiarity with different registers and speech genres, but competence varies from context to 386. Ibid at 8 [citations omitted]. 387. Marmor, “Chess,” supra note 360. Marmor has acknowledged that “the specific conventions we happened to have may matter to us, sometimes a great deal” (ibid at 169), and therefore “[t]here are political and moral values associated with rules of recognition” (ibid at 170), but this significance is not well captured by the chess metaphor. 388. See Bakhtin, “Speech Genres,” supra note 24 at 60 (“The wealth and diversity of speech genres are boundless because the various possibilities of human activity are inexhaustible, and because each sphere of activity contains an entire repertoire of speech genres”). 508 (2022) 59 OSGOODE HALL LAW JOURNAL context and can never be defined in a few rules of grammar.389 Further, legal communication does not occur in bits and bots, or in rigid binary codes (e.g., legal or illegal).390 From a Bakhtinian perspective, legal actors engage law and each other from their respective positions and perspectives, with their own registers and accents that may denote class, privilege, social distinction, and power-dynamics.391 Even within the law of a single national system, then, there is a world of difference between the law of the legal academic, the judge, the human rights and environmental activist, the tax inspector, the corporate lawyer, the politician, the law-clerk, the civil servant, the police interrogator, the public prosecutor, the member of an Indigenous community, the victim of gender violence, the conscientious objector, the anarchist, and all the various ways that these positions can intersect. In Bakhtinian terms, law inevitably operates in the midst of heteroglossia. Hence, legal language is not, like a chessboard, a neutral medium where the game takes place, but the very matter and material of discussion. In fact, the very analogy between law and games is questionable.392 Every time someone argues about legal norms in legal language, the arguments themselves are offered up for commendation or censure, and hence to be accepted, rejected, reinterpreted, and adjusted to the relevant context. In the process, law is susceptible to multiple projections, anticipations, interruptions, and echoes.393 389. Moreover, the idea of a basic grammar of law implies that there are neutral standards for distinguishing correct and incorrect uses of legal language. But no yardstick of legal-linguistic competence exists that could ever serve as referent across the board. Once we give up the idea of legal language as a matter of manipulating certain basic rules (or grammar), we may more easily accept that we can never know it all. See James Boyd White, “Legal Knowledge” (2002) 115 Harv L Rev 1396. 390. As, for instance, in Niklas Luhmann’s system-theoretical approach and in studies of Artificial Intelligence and algorithmic decision-making. See Law as a Social System, translated by Klaus A Ziegert (Oxford University Press, 2004). 391. See Pierre Bourdieu, Language and Symbolic Power (Harvard University Press, 1991). On the importance of sociolinguistics, see Goodrich, Legal Discourse, supra note 6. For an ethnographically-oriented critique, see Dell Hymes, “Speech and Language: On the Origins and Foundations of Inequality among Speakers” (1973) 102 Daedalus 59. From a feminist perspective, see Penelope Pether, “Critical Discourse Analysis, Rape Law and the Jury Instruction Simplification Project” (1999) 24 S Ill ULJ 53. 392. See Judith Shklar, Legalism: An Essay on Law, Morals and Politics (Harvard University Press, 1964) at 105-106. Shklar argues that “law and legal systems are not games but social institutions, and they do not exist in the social vacuum of a game....Unlike games [different legal orders] cannot be isolated, nor do they have simple purposes, a clear beginning, or an end” (ibid). See also Venzke, supra note 380, 352 at 354. Venzke rejects the analogy of legal interpretation to a game particularly “if that game is anything like chess” (ibid). 393. For various temporalities of law and judgment, see Etxabe, Tragic Judgment, supra note 270. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 509 Consequently, the activity of legal actors does indeed change the game, in a way that a chess player making a move does not. This takes us to our final and most important contention. If legal positivism had to be boiled down to a single idea, it would be that law necessarily forms a system. No idea galvanizes legal positivists more than the conviction that law forms a closed whole: a unitary and self-enclosed system that provides stability and coherence to what would otherwise be a world in disarray. In reality, however, this idea is necessitated only by the theory of legal positivism. Paraphrasing Wittgenstein, we might say that legal positivists are held captive by a picture.394 Allow me to flesh out this argument by addressing a typical line of defence of legal positivism against the charge that positive law is unable to determine the content of law without substantive evaluations.395 In an article that offers a rare glimpse into the heart of the matter, Sari Kisilevsky responds that unless we adopt the positivist position of value-independent social facts that can determine the content of law on their own, the resulting image “will bear little resemblance to anything we might call law.”396 In her view—a view we have seen reflected in every positivist author mentioned in this article—“If the moral significance of past law practices is…to be reevaluated every time someone says or does something new that might be of legal significance,”397 the resulting view would be so “stilted and fractured” that it would “shatter any notion of a unified system whose underlying values are expressed in law and can be uncovered so as to drive the system forward.”398 Thus baring the soul of positivism with crystal clarity, Kisilevsky affirms that “law also consists in a coherent system of rules or propositions that persist over time.”399 It is precisely “the systemic nature of law [that]…gives special weight to legal considerations.”400 394. As Wittgenstein wrote referring back to his earlier views on meaning produced as a calculus according to definite rules, “[a] picture held us captive. And we could not get outside of it, for it lay in our language and language seemed to repeat it to us inexorably” (supra note 363, s 115 [emphasis in original]). 395. See more recently, Mark Greenberg, “How Facts Make Law” in Scott Hershovitz, ed, Exploring Law’s Empire: The Jurisprudence of Ronald Dworkin (Oxford University Press, 2005) 225 at 225-26; Mark Greenberg, “Hartian Positivism and Normative Facts: How Facts Make Law II” in Hershovitz, ibid, 265. 396. “Against Legal Pragmatism: Greenberg and the Priority of the Moral” in Graham Hubbs & Douglas Lind, eds, Pragmatism, Law, and Language (Routledge, 2014) 269 at 277. 397. Ibid at 279. 398. Ibid at 280. 399. Ibid. 400. Ibid. 510 (2022) 59 OSGOODE HALL LAW JOURNAL Clearly, to say that a shift in the theoretical description of law would change law “as we know it” is not merely a description. Rather, it postulates the positivist idea of the legal system before and ahead of any alleged “empirical” investigation of “social facts” (as Raz would have it).401 And thus the “concept” of the legal system precedes any sort of law anyone can claim to have found.402 Yet, the real objection is not that legal positivists construct their own object of cognition, but that they seem unable to imagine law in any other way: Unable to free themselves from the image they have created, legal positivists remain captive to the limits of their own imagination, which is their language. Positivists conceive of language in an “abstract objectivist” manner as a system, a neutral medium, and propositional. Law is conceptualized in like manner in terms of rules, sentences, and structure: indeed, as an object of cognition to be apprehended and described as social fact in a detached way. This common view is projected throughout every dimension of law: legislation, statutory interpretation, adjudication, legal doctrine, jurisprudence, and of course, into the kind of law that is being taught, imagined, and reproduced in law schools. Positivists worry that without those qualities, law “as we know it” would cease to exist; that if law is not thought of as a system, then it would be impossible to hold it together. But what if, abandoning the presupposition that law is (and must be) a bounded whole, we were to start from the different presupposition that law is (or might be), as Margaret Davies suggests, “unlimited,” though “connected and relational”?403 Breaking free from this conception of language and law, I would like to propose a more generative metaphor. Relying on Bakhtin’s description of 401. Some legal positivists in the Kelsenian orbit are able to see this clearly. For instance, Uberto Scarpelli pointed out that “legal positivism...presupposes a prior political choice in favor of certain definition of law as the object of description.” Pintore & Jori, supra note 8 at 5. Scarpelli believed that legal positivism “does not lead to knowing the law as it is, law as fact, real law, but only knowing how the formal aspect of law can and must be when certain premises are accepted.” Enrico Pattaro, “Towards a Map of Legal Knowledge” in Pintore & Jori, supra note 8, 85 at 107. 402. As Kelsen argued on the basis of a Kantian epistemology, cognition thus “produce[s] the object of legal science by means of the...presupposition of what law must be like for the science of law to be possible” (Goodrich, Legal Discourse, supra note 6 at 37-38). This was the exact point of departure for Saussure as well: “[W]hat language must be like for a science of language to be possible” (ibid at 23 [emphasis in original]). Followers of HLA Hart may object that Hart’s project of “sociological jurisprudence” does not share the neo-Kantian transcendental premises of Kelsen, but, as we have seen in section VI, Hart’s “concept” is not an empirically verifiable fact, but a theoretical construct. 403. Davies, supra note 19. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 511 language as the uninterrupted stream of social interaction, I would analogize law to a stream of historical becoming that precedes us, propels us, and will naturally follow us, with no beginning and no end.404 Legal speakers enter the stream in which law comes to life and contribute to it as they move along. The vastness of the stream, its multiple undercurrents, and the lack of an immutable riverbed need not be a hindrance to feeling at ease and developing different degrees of familiarity with the diverse language(s) of law—in its plural manifestations, activity types, and genres. After all, a stream can be channelled, navigated, and traversed with different strokes and assorted vessels. This stream of law provides no solid floor, but as White has suggestively written, “[i]n this fluid world without turf or ground,” we can at least learn to swim.405 IX. CONCLUSIONS I began this critical article noting a disparity between the dialogical world around us and the dominant legal-theoretical language employed to describe it. Notably, some legal positivists continue to defend this account, not because it is a better or more apt account, but because it represents an ideal worth having. According to a variant of legal positivism labelled “ethical” (Tom Campbell) or “normative” (Jeremy Waldron), unless we are able to interpret and apply rules in a fairly non-controversial manner, the promise of the rule of law will be derailed—and with it, its capacity to curb the arbitrary exercise of power and coercion.406 According to this view, the acts of identifying rules and determining their content must be value-free because otherwise, legal interpretation would impermissibly confuse law-making and law-interpreting activities. Hence, the preference for positive rules that are “identifiable, discrete, clear, comprehensive, 404. Although with different association, the analogy between law and water exists also in the etymology of the archaic written form of the Chinese character of fà. See Alain Supiot, Governance by Numbers: The Making of a Legal Model of Allegiance, translated by Saskia Brown (Bloomsbury, 2017) at 63. See also Yan Lianke, Hard Like Water, translated by Carlos Rojas (Grove Press, 2021). 405. James Boyd White, “Law as Rhetoric, Rhetoric as Law: The Arts of Cultural and Communal Life” (1985) 52 U Chicago L Rev 684 at 695-96. 406. Tom Campbell, “Democratic Aspects of Ethical Positivism” in Tom D Campbell and Jeffrey Goldsworthy, eds, Judicial Power, Democracy and Legal Positivism (Ashgate, 2000) 3 [Campbell, “Ethical Positivism”]. See generally Campbell, Legal Theory, supra note 291; Waldron, supra note 292. For a good appraisal of the implied political philosophy, see Martin Krygier, “Ethical Positivism and the Liberalism of Fear” in T Campbell & J Goldsworthy, eds, supra note 406 59. 512 (2022) 59 OSGOODE HALL LAW JOURNAL consistent and applicable”407 in a value-neutral manner rests on a prior political aspiration, namely that governments function through the medium of rules that are capable of being identified and applied without recourse to contentious personal or group political presuppositions, beliefs, and commitments.408 In short, ethical (or normative) legal positivism is “wedded to the political significance of channeling governmental power through specific unambiguous conduct-governing mandatory rules…capable of being applied without recourse to contentious moral and political judgments.”409 In this article, I have sought to challenge not so much this ideal, as its viability—namely, that language understood in positivist terms could ever provide the assurances that the ideal requires. I have shown why even the most sophisticated software and computer-generated analysis is, without further intervention and critical evaluation, distorting, as it leads us astray from the actual task of legal interpretation. Strict adherence to corpus linguistics cannot serve the purposes for which positivists seek to enlist it. The sooner we come to terms with this fact, the sooner we will stop placing our hopes on “the most recent in a long history of similar Siren calls,”410 which continues to distract us from the real complications of legal judgment. Contrary to what might appear from those recent debates in law journals, interpretive struggles are rarely about individual words. Even when the discussion turns precisely to particular words or phrases (e.g., “vehicles” or “carrying a gun”), their meaning cannot be pulled up as if from a dictionary of standard occurrences. Consider that by the time a judge faces an interpretive question, the case is already languaged, filled with content, brought from the lived experiences of the parties into elaborated frames of discussion. A realistic account of this process must incorporate the rich tapestry of assumptions and prior texts that make it intelligible. Legal interpretation is thus a thick dialogical process: Words are interpreted with accompanying narratives and images and assessed from within the entire legal scene confronting the judge. The judge answers the call from a particular institutional position that is historically situated, elaborating on mutually affected and effecting language that is dialogized all the way through. I have argued that Bakhtin provides a more realistic appraisal of this living 407. Campbell, Legal Theory, supra 291 at 125. 408. This aspiration “depends upon the analytical thesis that law can be conceptually, argumentatively, and operationally separated from morality,” as well as upon “the sociological thesis that…laws are administered in a rule-deferential manner.” (ibid at 2). 409. Campbell, “Ethical Positivism,” supra note 406 at 4. 410. Francis J Mootz III, “Corpus Linguistics and Vico’s Lament: Against Vivisectional Jurisprudence” (2020) 20 Nevada LJ 845 at 846. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 513 stream of historical discourse. The Bakhtinian emphasis on utterances and the already uttered, the answering word and responsiveness, evaluative intonation, speech-genres, dialogical relations, double-voiced and internally dialogized discourse, et cetera, provides resources for richer and deeper engagements with the socially contested reality of legal language. Throughout, I have offered examples of legislation and statutory interpretation; of the language of doctrine and legal academics; of the language of judges; and of the language of normative discourse in general, as when we offer advice to a friend. Ethical positivists still insist that judges must “subjugat[e] their own moral beliefs about what would be substantively good law in favor of their moral commitment to obeying formally ‘good’ law.”411 Otherwise, there exists a risk that judges impose their own values upon society, which is contrary to the principle of separation of powers and the rule of law. The worry that judges might impose their own values on society is real, to be sure, but simply to exalt the role of formal rules does little to ease the anxiety. In fact, it can even exacerbate it, for when lofty ideals do not match the social perceptions of reality, the former may begin to corrode, leading to cynical and jaded views of what the rule of law is really all about. In contrast, bringing our theoretical language more closely into line with the dialogical world in which we live may prove salutary for those very same democratic aspirations. Admittedly, we may not want judges to substitute their own judgments for those of legislatures (although this way of putting things both overstates and simplifies the issue).412 Further, it may be conceded that we would not want judges to base their decisions on personal antipathies and animosities. Instead, we would hope that judges approach cases with open minds and care both for the parties and for society at large.413 But the central argument of this article is that one cannot know what “formally good law” is without fundamental evaluations, even though these evaluations need not be identical to the ones we 411. Campbell, Legal Theory, supra note 291 at 3. 412. Parliaments are not coterminous with democracy and examples abound of Parliaments disappointing or frustrating democratic aspirations. Alternatively, the judiciary serves important democratic functions against abuses of majoritarian politics and of the executive. On the judiciary as a democratic counter-power in a contemporary context where the traditional separation of powers has morphed into executive-driven forms of government, see Panu Minkkinen, “‘Enemies of the People’? The Judiciary and Claude Lefort’s ‘Savage Democracy’” in Matilda Arvidsson, Leila Brännström & Panu Minkkinen, eds, Constituent Power: Law, Popular Rule and Politics (Edinburgh University Press, 2020) 27. 413. For an excellent elaboration of the problem and options for responding, see Jeremy Webber, “A Judicial Ethic for a Pluralistic Age” in Omid A Payrow Shabani, ed, Multiculturalism and Law: A Critical Debate (University of Wales Press, 2007) 67. 514 (2022) 59 OSGOODE HALL LAW JOURNAL would adopt as “unencumbered” individuals. Judges operate within professional and institutional constraints that severely discipline their options and available languages, which they cannot always escape or control. Consider the current global phenomenon of judicial dialogues, which I alluded to in the beginning, where judges all over the world are citing each other’s words in mutual cross-references. Presumably, judges cite the words of other courts to increase their “persuasive authority,”414 but these words often deploy the strange ability to offer resistance and fight back, in ways that undermine the monological authority of the borrowing court.415 It is, as Bakhtin suggests, as if the words “put themselves in quotation marks against the will of the speaker.”416 Thus, the phenomenon of judicial dialogues is not simply to be seen as another unitary project of global law, because the sense of authority that emerges is not centripetal and one-directional, but diffuse and recursive. Finally, it may be further granted that judges should not ideally act in their own interests or on behalf of their supporters and ideological allies. Still, judges do not approach any legal matter as a blank slate, and preconceptions and fore-judgments are in fact part of any possible understanding.417 Moreover, judgement is also a matter of relationships and direction. Recall the earlier example of the Jewish friend seeking advice from the Catholic expert: It is precisely because the people involved are invested in their relationship that the Catholic friend remains personally responsible for any advice given.418 Hence to insert oneself in the situation doesn’t invalidate judgment; it makes it answerable.419 The example is illustrative for another reason: The adequacy of the advice cannot be assessed against a system of norms externally or cognitively 414. Patrick Glenn, “Persuasive Authority” (1987) 32 McGill LJ 261. 415. An interesting example occurs in the Hirst case cited above where a British judge cites the supporting opinion of the Canadian Court of Appeal’s upholding of the disenfranchisement of prisoners. The latter decision was later reversed by the Supreme Court of Canada, which was eventually picked up by the European Court of Human Rights and used as an argument, in refracted manner, against the UK legislation. See Hirst, supra note 214 at paras 35-36. 416. Bakhtin, “Discourse,” supra note 1 at 294. 417. This is the main tenet of the hermeneutical tradition of Gadamer, Ricoeur, Taylor, MacIntyre, and others. See also Jeanne Gaakeer, Judging from Experience: Law, Praxis, Humanities (Edinburgh University Press, 2019). 418. I do not mean to suggest that the relationship between judges and society is likewise one of “friendship,” or that we should simply trust judges to do their job. Vigilance is certainly constantly required, but the object of vigilance should not be aimed primarily at whether a particular judge has followed some formal rules or text, abstractly considered. 419. See Jennifer Nedelsky, “Embodied Diversity and the Challenges to Law” (1997) 42 McGill LJ 91. ETXABE, THE DIALOGICAL LANGUAGE OF LAW 515 apprehended. Rather, it requires a holistic assessment of dispositions, sensibilities, and responsiveness to the situation. For judges, too, a similar assessment could be made. We may ask, for example, how does the judge build relationships with the parties? Does the judge show openness and receptivity to the normative worldviews in dispute? Does the judge represent their views adequately? How is normative disagreement addressed? Further, how are jurisdiction and the authority of the court established? What symbols, narratives, and imagery are built into the scene? What is the level of engagement with prior discourses and narratives? Does the judge struggle with them or accept them blindly? What responses does the judge anticipate and which ones does the judgment invite for the future? Fundamentally, we might ask: “Is this a conversation in which democracy begins”—or one in which it ends?420 In conclusion, to argue as I do that ascertaining the law requires normative evaluations from judges and critics alike is not to impermissibly insert subjective elements into legal adjudication. Rather, it is to open up the space of judgment to the dialogical language of law. Law is not a text that a sender (i.e., the legislature) creates for others to passively apply in a conveyor belt manner. Law is living and historically unfolding in an ongoing, inter-orientated manner. Similarly, legal judgment—whether done by judges or not—is not a subjective construction of the individual mind, but an interactive, embedded, and responsive activity whereby our entire being is at stake, whether writ large or writ small. Mediated by our perceptions of the obligations, demands, possibilities, and constraints of the situation, we respond with our being to the event of judgment and remain answerable to it. If legal judgement is to respond to our most cherished democratic aspirations, then it must be, like language itself, dialogical. 420. See White, Justice as Translation, supra note 187 at 113.