EL CÓMPUTO SILÁBICO

COUNTING SYLLABLES

EDOARDO ESPOSITO
Università degli Studi di Milano

Resumen: En un libro reciente, Meter in Poetry. A New

Theory, los autores Nigel Fabb y Morris Halle, han propuesto, ba-
sándose en la métrica generativa, un sistema de análisis del verso
que se puede aplicar, con las oportunas adaptaciones, a todo ám-
bito lingüístico y a la poesía de todos los tiempos. La mayor parte
de las páginas se dedica en efecto al análisis de textos que van del
griego antiguo al inglés moderno y a todos los otros “meters of the
world”, mientras que en el primer capítulo se enfocan la teoría y
las modalidades de análisis. Es aquí donde justamente se patentiza
la fragilidad del planteamiento (cada regla admite un sinnúmero de
excepciones y también su contrario) y sobre todo la inconsistencia
de un sistema que pretende estudiar el metro prescindiendo del rit-
mo y de la semántica, como si un verso fuera tan sólo un conjunto
de signos gráficos.

Palabras clave: metro, ritmo, poesía, teoría de la literatura,

teoría generativa, sistemas métricos del mundo.

Abstract: In their Meter in Poetry: A New Theory, a recent

book on generative metrics, Nigel Fabb and Morris Halle pro-
pose a method of parsing verses that applies to every linguistic
area and to the poetry of every time. Its main part is dedicated to
parsing texts that go from ancient Greek to modern English and to

—43—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

all the other “meters of the world”, while the related theory and
rules are set in the first chapter. Here lies, however, the weakness
of the system, as every rule presents lots of exceptions and even
comes to admit of its contrary. Most of all, the method seems fal-
lacious in that it pretends to study meter as detached from rhythm
and semantics, as if the line were only a sum of graphic signs.

Key words: meter, rhythm, poetry, literary theory, generative

theory, meters of the world.

—44—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

I N an article that appeared recently in this review1, reference

was made to the study Meter in Poetry. A New Theory2. Be-
yond featuring two authors as renowned as Nigel Fabb and
Morris Halle, the book originates in the “generative” context,
which is nowadays one of the few to contribute to the field of
literary theory after its decline in the last decades of the twenti-
eth century. One could argue whether this is actually theory or
whether Fabb and Halle have, in fact, developed a formulary, a
series of methodological indications aimed at identifying what
they call a «well-formed line». The authors, however, do not lin-
ger on philosophical questions; and while it is true of any theory
of poetic form that it implies a theory of poetry, let the readers
be reassured: they will be spared both.
Furthermore, only the first chapter of the book “A theory of
poetic meter” raises basic issues in this regard, such as the rela-
tionship between meter and rhythm, or a possible definition of
“verse”. The remaining ten chapters that make up the book are
dedicated to the analysis of different verse forms –or, rather, to
the application of the analytical rules presented in the first chap-
ter. These verse forms belong to different traditions and ages,
from Sanskrit, through ancient Greek, Arabic, old and modern
English, Spanish, French to any other “Meter[s] of the world”
(ch. 10), although these last are only touched on. The ambition
of this book is to define universally valid principles for the as-
sessment of «metrical verse, which is the most widely used kind
of poetry and is also the subject of this book» (p. 3).
1
See DUFFELL, Martin J.: “The principles of free verse in English”, Rhythmica. Revista
española de métrica comparada, 2010, 8, pp. 7-35. Thanks to Sara Sullam for this
translation.
2
FABB, Nigel & HALLE, Morris: Meter in Poetry: A New Theory. Cambridge: Cam-
bridge University Press, 2008.

—45—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

Carlos Piera, a highly regarded member of the scientific com-

mittee of «Rhythmica», has contributed to the research with
the chapter “Southern Romance”. Since Meter in Poetry is a
remarkable undertaking, with which future studies in the field
will necessarily deal, I think it is worthwhile considering once
more its guiding assumptions before they become the basis for
future reflections. I will not dwell, however, on specific textual
analysis, which have already been carefully looked into by other
reviewers3; I will limit myself to recalling that on this very point
many objections have been made, since the applicability of the
principles elaborated by Fabb and Halle for English poetry actu-
ally requires a radical revision to be made every time a differ-
ent language and cultural tradition is considered. This strongly
impairs the theoretical value of their proposal, which, in fact, is
puzzling from its very distinction between «strict» and «loose
meters» in English poetry.
Meter in Poetry has generated attention and discussion, which
is inevitable for a work in which –as already pointed out by other
reviewers– «Ambitions for cognitive relevance and universality
are set high»4. However, reviewers have not focussed on its pre-
mises, partly due to the fact that debate has taken place mainly
within a generative context –in this case, a generativism not
excessively concerned with defining itself and its specificities.
Hayes, in fact, warns that «scholars new to generative metrics
would be well advised to do some background reading before
taking on the challenging proposals presented in this work» (p.
2520), while Riad observes that Fabb and Halle «do not seriously
challenge any other brand of generative metrics» (p. 542).
A first, in fact preliminary, reason for dissent –a methodologi-
cal rather than critical dissent– is the fact that in the title of their
book Fabb and Halle do not present their theoretical proposal as
3
See the reviews by Paul KIPARSKY in Language, December, 2009, 85, 4, pp. 923-930;
Tomas RIAD in Phonology, 2010, 27, 3 pp. 542-551; Bruce HAYES in Lingua, 2010,
120, pp. 2515-2521.
4
RIAD, p. 242; KIPARSKY has defined the book a «major event» (p. 923) and has ob-
served that «the strengths of this book are the clear exposition, and the application
of a carefully worked out, phonologically grounded theory to interesting mate-
rial representing the whole typological spectrum». His conclusion, though, reads:
«The results are disappointing» (p. 929).

—46—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

“generative”. Had they done it, readers unaware –if not of the
perspective of that critical context, of its language and critical
tools– would have been better equipped to read the book, they
would have more eagerly accepted the substitution of concrete
syllables and words through abstract symbols (“asterisks”) and
could have more easily understood the mechanisms of “projec-
tion” and “grouping” transforming a line of verse into a not less
abstract “grid”; and perhaps they would not have been startled
at reading statements such as: «John Donne uses non-projection
very extensively» (p. 60). Moreover, it seems to me that speak-
ing so generically of a “new theory” –instead of a “new gen-
erative theory”– implies a substantial disregard for what is not
“generative”, especially if no mention is made (and this is the
case) of different critical traditions, or if no effort is made (and
this is the case) to present one’s positions and assumptions and
to justify them, to compare them with the achievements of other
theoretical proposals and, more generally, to discuss them. This
inevitably causes unpleasant lapses; so, for example, does Riad
feel obliged to notice that one is faced with a «fairly novel ap-
proach to Greek metrics» that turns out be «clearly at odds with
traditional analysis and also with phonological fact». Kiparsky,
on the other hand, comments: «This intricate theory is devel-
oped with precision, but with little justification», and adds that:
«The daunting task of assessing the theory is left to the reader»
(p. 925).
I’d like to start, though, from the statement according to which
«What distinguishes all poetry from prose is that poetry is made
up of lines (verses)», immediately following the definition of
poetry as «a form of verbal art» (p. 1). One is not shocked by
the fact that, in order to define poetry and to distinguish it from
prose, the authors resort to a purely formal criterion; rather, what
I find puzzling is the fact that they invoke writing (which is al-
ready pretty rough, since it has already been objected that even
gravestones can be written “in lines”) without considering that
writing simply gives visibility to and transmits something origi-
nating as a verbal message, with all the features this implies: first
of all, it takes place in time (it has a “duration”: the end of a line
of verse is simply the end of the melodic-rhythmic line which is

—47—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

consubstantial to it; and it signals a new beginning, delimited in

tone by a pause and marked, in writing, by a fresh line). Secon-
dly, it is characterized by precise sonorities, by precise intensi-
ve or iterative phenomena. These phenomena are perceived and
appreciated by readers even when they read the text silently, in
their interiority; for a word is, in our understanding, first and
foremost a sound, and each grapheme is perceptibly and inevi-
tably associated to a sound and to a duration (and we will not
be dealing with semantics and syntax here, not because it is im-
possible, but rather because Fabb and Halle do not, and doing
it would mean entering an even more polemical discussion).
Our authors quote Tomaševskij, who said that, in language,
«the phenomenon of verse itself does not exist»; Fabb and Halle
seem to forget the radically different positions that emerged in
the critical context to which Tomaševskij belonged. Of course,
the phenomenon of verse itself does not exist. In fact, a verse
“exists” only if it is accompanied by other verses and it beco-
mes part of a series: only in this case does its entity pass from
potentiality to act and becomes recognizable as such. This, how-
ever, does not imply denying that even ordinary language can be
“interwoven” with verses, nor does it lead to stating that «there
is a well-founded distinction between texts divided into lines
and texts not divided into lines» (ibidem): that distinction does
not exist, it is a mere convention –as anyone who is acquainted
with ancient codes, in which verses were written consecutively,
already knows; while it is undoubtedly true that if the Divine
Comedy or Paradise Lost were written consecutively, this would
not prevent the reader from correctly deciphering its rhythmic
nature. Hopkins was right when he said that the underlying prin-
ciple of verse is parallelism, while the contrary cannot be said:
a verse simply provides evidence for the rhythmical construct
realized by the poet.
I’d like to make it clear, once again, before moving on: be-
fore being written, a verse is a sequence of sounds, or rather, a
series of linguistically organized phonemes, and, as in any other
written text, it records a word which has been –or still has to
be– communicated. As we’ve been taught, therefore, it is a re-
presentation of something with an oral nature, the features or

—48—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

functionality of which cannot be assessed regardless of sono-

rities making it up and of the time necessary for them to be ex-
pressed. If we do not keep this into account, we run the risk of
mistaking what is in movement with its inert traces:
Les spécialistes du rhythme poétique se noyaient dans les vers, les
divisant en syllables, en mesures et essayaient de trouver les lois du
rhythme dans cette analyse. En fait, toutes ces mesures et syllables
existent non pas en elles-mêmes, mais comme résulat d’un certain
mouvement rythmique. Elles ne peuvent donner que des indications
sur ce mouvement rythmique dont elles résultent. Le mouvement ryth-
mique est antérieur au vers. On ne peut pas comprendre le rythme à
partir de la ligne des vers; au contraire, on comprendra le vers à partir
du mouvement rythmique5.

Fabb and Halle, though, seem to be content to work with

“traces”, on a form that they consider as static, fixed on paper
and defined not in temporal terms, but in spatial ones. Which,
after all, Halle had already done when he founded the so-called
generative metrics with Samuel Jay Keyser, substituting the tra-
ditional term syllable with position –that is, with a spatial term6.
Fabb and Halle go back to syllables and verse length, precisely
measured by the number of syllables (p. 3)7. It should not be
forgotten, however, that neither syllable nor verse “length” are
actually unambiguously defined. What I find even more serious,
though, is that, in the analytical hypothesis that we are consider-
ing, this kind of reality is immediately put aside in favour of an
even more abstract representation of such reality (the asterisks
we have already mentioned), in which any link with the lan-
guage making up the verse and with the rhythm shaped by the
verse is broken once and for all.
This could still make sense if, as in algebraic representations,
there was a precise correspondence between that which repre-
sents and that which is represented, and if, therefore, any “trace”
5
BRIK, Osip: “Ritm i sintaksis”, Novyj Lef, 1927, 3-6; in Théorie de la littérature,
textes des Formalistes russes réunis, présentés et traduits par Tzvetan Todorov,
préface de Roman Jakobson. Paris: Editions du Seuil, 1965, pp. 143-153: 144.
6
See HALLE, Morris & KEYSER, Samuel J.: “Chaucer and the Study of Prosody”. Col-
lege English, 1966, XXVIII, pp. 187-219.
7
«The core of the theory (...) is about the measuring of line length and the grouping
of units projected from syllables» (RIAD).

—49—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

objectively referred to an element of a given verse. However, as

always happens in these cases, the question concerns what can
be defined as “objective”, that is, the way in which we can elim-
inate our interpretation from the act of deciding what has to be
represented and how it is to be done. Here one is faced with an
issue which is crucial and even more so, in my opinion8, since it
is left unsolved the relationship with the phonological reality of
verse (of language), which, under certain aspects, is not made
explicit enough, and otherwise seems to take for granted that
the graphic transcription of language is absolutely “objective”
(see above) and is totally consistent with its verbal features.
The problem is that our authors seem to be totally uninter-
ested in rhythm, and they know it: «In this respect our approach
departs radically from most other approaches to meter, as these
have been focused almost exclusively on rhythm». In their per-
spective “the rhythm is a by-product of the way line length is
restricted»9 (p. 3) and one should rather conform to meter, since
it is to meter that the reality of verse aims. But what is “meter”?
In a previous study10, Nigel Fabb hypothesized «some special-
ised kind of ‘metrical cognition’ as one of the types of linguistic
cognition» (p. 13). While this kind of metrical cognition still
lacks a precise definition, the rules and conditions controlling
it have by now been made clear: «A meter is a set of rules and
conditions» (p. 93). In fact, a major part of the book is dedi-
cated to the development of such rules11, to the identification
8
But see also RIAD, who, after stating that «It is everybody’s intuition that there is a
connection between phonology and metre» and after underlining the relevance
of phonological factors in FABB and HALLE’S observations, notices that, having
to deal with issues raised by specific metrical traditions, «the connection is less
clear. Phonological notions employed in metrical conditions are simply stated, or
taken as self-evidently given from the language» (p. 547). HAYES, on other hand,
wonders: «does metrics includes instances where the data pattern is orderly in
underlying form but not at the surface?» (p. 2517) and he points out, for instance,
that «a tension can be seen between FH’s use of the “strongest position” concept
–namely, as defining the site of special rule application– and more traditional con-
cepts that relate more intuitively to metrical strength» (p. 2518).
9
KIPARSKY, p. 924. The author underlines this aspect of the text, proposing an analyti-
cal discussion of it.
10
FABB, Nigel: Language and Literary Structure. Cambridge: Cambridge University
Press, 2002.
11
Already partly elaborated in FABB, cit..

—50—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

of the mechanisms best suited to make verse analysis (of any

verse) reliable and objective, not depending on interpretation
and aimed at highlighting the “metrical grid” underlying a cer-
tain verse measure, the pattern, which «though non pronounced,
determines the perception of a sequence of syllables as a line of
metrical verse, rather than as an ordinary bit of prose» (p. 11).
In fact, in such a perspective «a verse line is well formed metri-
cally if and only if its grid is well formed» (Ibidem. Therefore, if
reality does not match the criteria with which we have decided
to analyse it, well, then so much the worse for reality...)12.
Let us follow, however, the process allowing us to construct
the grid and which, as we said, reduces the linguistic entity of
the verse to a series of abstract elements which take the place
of syllables and which are grouped according to the «iterative
rule»,13 from left to right or from right to left, following a bi-
nary or ternary principle; the last group of syllables can be left
incomplete.
Here is a verse by Keats; let’s follow its “transformation”:
Ever let the Fancy roam,
* * * * * * *
(* * (* * (* * (*

And here is Byron:

For the Angel of Death spread his wings on the blast
* ** * * * * * * * * *
)* * *) * * *) * * * ) * * *)

Those who are acquainted with generative metrics already

know that this process, explained in the Fabb’s Language and
Literary Structure, implies several further steps, which cannot
be analysed in detail here. It is there that we have to look for the
explanations that the author makes no effort to provide in Meter
in Poetry; we thus find a series of «projection and non-projec-
tion rules» (LLS, p. 8) which applied, for instance to a sonnet by
12
KIPARSKY similarly observes that FABB and HALLE «blame Hopkins for violating
their own wrong rules» (p. 928).
13
On this point see IDSARDI, William: The Computation of Prosody. PhD Dissertation,
MIT, 1992.

—51—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

Browning with its syllabic variability, allow for it transforma-

tion «into a metrical representation in which there are exactly
ten syllables in every line» (Ibidem).
Traditional metrics yielded similar results assessing the leng-
th or shortness of vowels and resorting to “metrical figures” such
as synalepha; those results, however, often relied on interpreta-
tion, while, indeed, getting them by means of objectively appli-
cable rules or, at least diminishing the number of factors usually
taken into account, would be a relevant fact from a scientific
point of view. Unfortunately, this does not seem to be the case
with Fabb, for, while certain lines of the sonnet “project” exactly
10 asterisks:
Remember, never to the hill or plain,
* * * * * * * * * *

some of them, inevitably, project 11 or 12; one has therefore to

establish, for example, that the last syllable of the verse «which
is unstressed or weak in stress and which comes after a strongly
stressed syllable» (p. 9) must not be projected.
That thou dost love me. Though the word repeated
* * * * * * * * * * Δ

Perplexity does not arise from the necessity of a new symbol

(one of “non-projection”: Δ) or from the need to take into ac-
count an exception beside the rule, but rather from the fact that
identifying unstressed, weak in stress, strongly stressed syllable
will not always be as easy as in the abovementioned example. Of
course, it can be in most cases: still, wouldn’t we end up relying
on the very phonological reality that the new theory claims not
to consider? And what, then, is the novelty with respect to the
traditional notion of meter? There, too, one relies on accents to
decide whether to “count” a syllable or not. In Italian, for instan-
ce, one counts only one syllable after the last accent of the line,
even though there might be two or three of them; and there is
also the possibility to have a “virtual projection”, since after the
last accent one usually counts a further syllable even if the line
is concluded. And the abovementioned «non-projection rule»
—52—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

(a) is not the only one; for it is followed by a «non-projection

rule» (b) and a «non-projection rule » (c). The former reads:
Optionally: do not project a syllable which ends on a vowel, when
that syllable precedes a syllable which begins on a vowel. (p. 9)

And the latter:

Optionally: do not project a syllable which has as its nucleus one
of the following sonorant consonants: [l], [r], [m] or [n], or which has
as its nucleus the weak vowel [ə] followed by one of these sounds.
(p. 10)

These rules are to be reconsidered keeping into account the

phonological system of every different language; in the Italian
system, which I have just mentioned, rule b is provided for by
synalepha while rule c has no evidence. The problem is that, in
such cases, we should not speak of rules but rather of possibili-
ties, which can apply or not: for they are optional and it is up to
the interpreter to decide whether they apply or not. Where, then,
has “objectivity” gone? If we want to change a system, we have
to come up with tools which turn out to be more efficient than
older ones, less subject to the evaluation of the single person.
In the sonnet by Browning quoted by Fabb, all accents fall on
an “even position”, but, given the optional character of “non-
projection rules”, in other cases it will always be possible to
conform reality to quod demonstrandum erat. Fabb himself im-
plicitly admits it when he states (p. 13) that «it is true of almost
iambic pentameter lines that they have ten projected syllables.
That is, we can do to almost every line of iambic pentameter
what we did to the Browning poem, with the same results». The
fact that this does not always happen, but rather almost always
(emphasis mine) does not seem to worry Mr. Fabb; after all, any
work made of pentameters is, of course, made of pentameters.
Let’s get back to Fabb’s new book, though. It should contain
clear indication as to the reasons for the adoption of certain pro-
cesses –it does not, though. One wonders, for instance, why syl-
lables are “projected”, and not single phonemes; why mention is
made only of binary and ternary principles; why the grouping of

—53—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

syllables can start from left or right likewise and why each pro-
jected “gridline” can change what held of the previous gridline
level14. Since all of this is lacking, what we read on p. 5 «the me-
ter controls primarily the number of groups in the line, and only
secondarily the number of syllables» sounds definitely overas-
sertive, inasmuch as it subreptitiously mistakes «the meter» with
what is simply Fabb and Halle’s notion of it. This is precisely
the methodological stance of the two authors: while they do not
provide any reasoning or reflection generating curiosity in their
readers, let alone trust, they force them to mechanically follow
the process. One often gets the impression that rules presented
as new –and which are each time subject to variations– are in-
spired by linguistic facts and traditional principles (which are
not defined as such, of course: see, for example, the importance
attached to accents). «As you like it», we might say, borrowing
the words of someone who knew metric problems very well.
Moreover it has to be said that even when some reasoning
is present at all, it turns out to be hasty and hard to share. For
instance, dealing with enjambment, separating not only syntac-
tical phrases but occurring also –albeit rarely– «in the middle
of the word», Fabb and Halle conclude “the fact that lines are
sequences of syllables, rather than of words or phrases» (p. 10).
They don’t seem to realize that: 1) they are founding their rule
on exceptions, since enjambment is not a rule, but rather an
exception in poetic discourse (and it is precisely why it has a
name); 2) that an enjambment «in the middle of the word» is an
exception to the exception, since evidence is very scarce in this
14
I think that in the description provided by KIPARSKY for these operations, the ran-
domness of the conditions controlling those very same operations becomes clear:
«For each level in each meter, parameters determine the direction of scansion,
the orientation of the parentheses, whether intervals are binary or ternary, and
whether the parse begins at the edge, or one or two asteriks in. Additional “riders”
specify whether the resulting groups can be, or must be, incomplete at one edge,
and whether some syllables can or must remain ungrouped. Before grid construc-
tion begins, brackets may be inserted by rules sensitive to weight, linear context,
or alliteration. At any point in the derivation, rules may delete asterisk and pa-
rentheses, apparently at any gridlevel, in contexts defined either hierarchically
by asterisks and parentheses, or linearly by the weight or stress of neighboring
syllables. These deletion processes allow groupings of any lenght to be formed».
Kiparsky concludes later: «Hybrid system of rules and constraints have the major
disadvantage that they lead to difficulties with managing their interaction, and to
undesirable duplication» (pp. 924 -925).

—54—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

case; 3) that verse is not a single unit, but it exists, as pointed

out above, because it is part of a sequence; a sequence in which
the fact that a verse ends with an enjambment is definitely not
without prosodic, intonational and rhythmical consequences.
Unfortunately, the abstraction from the reality of the data to
be examined easily inspires fallacious interpretations; and such
an operation is a bad way of referring to the notions of genera-
tivism. Referring precisely to Chomsky (p.12) the authors claim
that «the rules of meter (...) generate abstract elements that are
not directly present in the acoustic speech signal»; what I find
puzzling, however, is the fact that they resort to the term ele-
ments for an abstract model which is apparently not «directly
present etc.»: if it is recognizable, it is because its rules oper-
ate –and are therefore present– in a «speech signal», even if, of
course, it is not one of its “elements”. And I’d like to add that it
has been acknowledged that it identifies with a musical and tem-
poral principle, created by the linguistic string itself or to which
the string adjusts (we will return on this point later).
Not less puzzling is the discovery that «words and syllables
are not overtly present in the acoustic speech signal of the utter-
ance; they are constructed by speakers and hearers alike by vir-
tue of their knowledge of the language» (p. 11), for it is at least
since Saussure’s time that it has been acknowledged that spoken
language is a phonic chain in which pauses are motivated only
by needs of breathing, intonation or emotional stressing. I’d like
to add the need (often a strong one) for semantic disambiguation,
but all this, while causing alterations in metrics, does not obvi-
ously imply that the model of reference should be questioned;
moreover, one could argue over the fact that the capacity to rec-
ognize all this is «part of the human capacity for language» (p.
12) rather that something that is acquired culturally, and to this
regard I refer to Franco Brioschi’s work, which is unfortunately
written in Italian15.
The main problem, however, is that the conditions and restric-
tions proposed for the construction of the “grid”, intended to tell
15
See BRIOSCHI, Franco: La mappa dell’impero. Problemi di teoria della letteratura.
Milano: Il Saggiatore, 1983 (in particular: “Il lettore e il testo poetico”, pp. 57-
114).

—55—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

us whether a line of verse is “well-formed”, do not really gener-

ate a clear series of what it is possible –or not possible– to do,
because the series can be changed and reset each time something
turns out to be an exception. This, of course, is totally compatible
with the elaboration of a new theory, which, in fact, should be
able to account for the highest number of cases: problems arise
when what the theory justifies actually applies only to a limited
number of cases and, once a rule is established, one immediately
has to establish the possibility of its contrary: «in strict meters
the grouping of the syllables on Gridline 0 determines the place-
ment of the maxima, whereas in loose meters it is the placement
of the maxima that determines the grouping» (p. 33). The fact
is that there is no rule to tell “strict” metres from “loose” ones;
or, rather, to determine it one has to go back to reading the lines
and the words that make them up, and the “grid” ends up being
a schematization of what we already knew16.
That a process of simplification should be intrinsic to gen-
erative analytical modes is widely acknowledged, and so is the
tension to the future inspiring research of this kind: «at this time,
next to nothing is known about the neurophysiology of syntax
and semantics, and about how actual speakers compute the well-
formedness of the sentences they produce. This problem may
be solved in the future, as we get a better understanding of how
the human brain and mind work» (p. 9). And yet, for the future
to be correctly scanned, from a methodological point of view
the instrument through which we observe it should be fit to the
reality we want to analyse, the very same reality from which the
canons for observation should be drawn. In the meantime, we
could more modestly be satisfied with understanding in which
sense the grid –which we are advised to build and which ends up
being different for every “metrical tradition”– can grant access
to the concrete complexity of a line of verse.
16
HAYES resorts to similar arguments when he writes: «the theory allows languages to
vary a great deal in their constraints. For instance, the theory allows constraints
that require that strong positions be filled by a heavy syllable (p. 164), but also
constraints that require that weak positions be filled by a heavy (p. 229)». While
he recognizes that «This is not an objection in itself, since these constraints, after
all, are doing real work in the analysis», he observes that «the sheer variety of pos-
sible constraints makes it hard to assess the restrictiveness of the overall theory»,
and hopes the authors will work on «a computation of the output typology of their
theory» (p. 2517).

—56—
LA ESCANSIÓN SILÁBICA Rhythmica, IX, 2011

The authors are convinced that grids «allow us to distinguish

metrically well-formed lines from other syllable sequences»
(ibidem); I suppose that in this way a machine may generate a
“well-formed line”, though I doubt that anyone who has a mu-
sical ear and a good literary education could have some gain
considering the effort needed to understand the mechanism. I
think it would be more relevant to understand why poetry can
also include “bad-formed lines” without any major damage, and
why an iambic pentameter is recognized as such in spite of its
“irregularities” and the “bad manners” with which Shakespeare
and other treated it. Any “grid” will end up discarding a substan-
tial part of the verses that make up literary history, because poets
are not really interested in being “in order”.
Meter is a «set of rules and conditions» only inasmuch as
these rules tell us how to measure a line of verse, what makes
it comparable with other lines, or what, potentially, makes it the
“same” as other lines. Sameness is never a given fact; it is only
through comparison that rhythm makes a pentameter the same
as another and an endecasillabo the same as another.
Fabb and Halle’s book does not fail to make a reference to
music, acknowledging its similarity with verse; however, the au-
thors do not go beyond stating the mere fact (which, apparently,
is considered discriminating) that music is measured by «time
intervals» (p. 36) which can also include silence, while poetry is
a matter of syllables which can never be reduced to silence. And
yet poets have often said that their poetry is born out of the de-
sire to «dress» with words a kind of interior music17. It would be
enough to admit that the scheme the consistency of which Fabb
and Halle look for –believing that they will ascertain its speci-
ficity– is simply the scheme of time. With the variations and, in
certain cases, the irregularities of their concrete rhythm (which
17
See for example Paul VALÉRY: «Si donc l’on m’interroge (…) je réponds que je n’ai
pas voulu dire, mais voulu faire, et que ce fut l’intention de faire qui a voulu ce
que j’ai dit. Quant au “Cimetière Marin”, cette intention ne fut d’abord qu’une
figure rythmique vide, ou remplie de syllabes vaines, qui me vint obséder quelque
temps. J’observai que cette figure était décasyllabique, et je me fis quelques réflec-
tions sur ce type fort peu employé dans la poésie moderne; il me semblait pauvre
et monotone» (in COHEN, Gustave: Essai d’explication du Cimetière Marin, suivi
d’une glose analogue sur La jeune Parque precédée d’un avant-propos de Paul
Valéry au sujet du Cimetière Marin. Paris: Gallimard, 1946, p. 22).

—57—
Rhythmica, IX, 2011 EDOARDO ESPOSITO

is not controlled by a chronometer but rather by our psycho-

physical perception), verses simply scan segments of regularity
of time. Being a fairly perceptible reality, close enough to the
“time intervals” of music, the division into syllables has tradi-
tionally been used for line-measuring, and we might keep doing
so; though, if we want to get closer to a scientific definition of
a line of verse, of what it is and how it works, we do not need
to build abstract grids referring back to the syllable. Rather, we
should look at the way in which the syllable fills time intervals.
With the notion of time we can account for both realities, with-
out having to resort to a mysterious “cognition”: entia non sunt
multiplicanda praeter necessitatem.

—58—