Tasting notes

Tasting notes: A corpus-based study of olive oil and wine tasting discourse Belén López Arroyo & Lucía Sanz-Valdivieso University of Valladolid (Spain) mariabelen.lopez@uva.es, lucia.sanz.valdivieso@uva.es Abstract Tasting notes are professional texts used in different specialized contexts with the purpose of organizing the taster’s sensory perceptions into attributes. There have been multiple studies focusing on the linguistic features of tasting notes, from their rhetorical structure to their use of metaphors; however, they have never been analysed using a combination of different, but complementary, linguistic perspectives, genre, and register. Our methodology, by employing these approaches, will outline comprehensively their features. In this paper, we analyse the genre and register features of tasting notes in two different specialized languages using a corpus to find out whether there is a disciplinary variation or not. Additionally, we will describe, classify, and contrast the way information is organized at different levels of analysis. Our results will be useful for scholars of genre, register and discourse studies, and for experts and technical writers in the olive oil and wine sectors. Key words: LSP genre, register, rhetoric, terminology, phraseology, corpus linguistics. Resumen Estudio del discurso de la cata en un corpus de fichas de cata de aceite de oliva y vino Las fichas de cata son textos profesionales que se usan en diferentes contextos especializados con el propósito de organizar las percepciones sensoriales del catador en atributos. Existen múltiples estudios que analizan aspectos lingüísticos de las fichas de cata, desde la estructura retórica hasta el uso de metáforas en ellas; sin embargo, no hay estudios que hayan combinado perspectivas diferentes, aunque complementarias, como son las del género y el registro. En el presente artículo, analizamos los rasgos de género y registro en un Ibérica 43 (2022): 205-234 ISSN: 1139-7241 / e-ISSN: 2340-2784 205 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO corpus de fichas de cata con dos lenguajes especializados, el del vino y el del aceite, con el fin de descubrir si presentan variación disciplinar. Además, describiremos, clasificaremos y compararemos la forma en que se organiza la información de las fichas de cata en diferentes niveles de análisis. Nuestros resultados serán de utilidad para lingüistas centrados en estudios discursivos y de registro, expertos en los campos objeto de estudio y escritores técnicos. Palabras clave: género especializado, registro, retórica, terminología, fraseología, lingüística del corpus. 1. Introduction Tasting notes (TNs1) are professional texts used in the wine and olive oil sector with the purpose of organizing the sensory perceptions for each product into attributes (Diederich, 2015: 36). They are short standardized texts in which products are described, evaluated and often rated along a scale to record the different organoleptic features or components of wine or olive oil. TNs can be analysed by focusing on different linguistic perspectives in order to comprehensively describe their features, from the point of view of both the register and the genre to which they belong. As Biber and Conrad (2019: 76) suggest, both perspectives are complementary since, “although the register and genre perspectives analyze different types of language features, it is often useful to add an analysis of genre features when undertaking a register analysis, in order to describe the text variety more fully.” Even though wine tasting notes (WTNs) have been extensively analysed and studied from different points of view (Caballero, 2007, 2017; López-Arroyo & Roberts, 2014, 2016; Paradis, 2010; Suárez Toste, 2017, among others), none of them have approached these texts using both register and genre analysis. At the same time, tasting notes are also used in other specialized contexts, such as the olive oil sector, but few studies exist on the description of olive oil tasting notes (OTNs) apart from some lexical and lexicographic analyses of language use (Montoro del Arco, 2012; Roldán Vendrell & Fernández Domínguez, 2012; Montoro del Arco & Roldán Vendrell, 2013a, 2013b; Roldán Vendrell 2007, 2010, 2013a, 2013b; Santa María, 2013). We are unaware of any studies that combine register and genre analysis in TNs or explore these factors regarding wine and olive oil tasting notes. In this paper, we will thus compare WTNs and OTNs to see whether there is 206 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE variation (Parodi, 2013) or a significant number of similarities between the TNs in these two different specialized professional discourses. Our hypothesis is that the structure of the TNs does not vary from one context to the other. Particularly, we aim to: 1. Describe to what extent WTNs and OTNs share the same register. 2. Analyse if they have the same function and are linked to the same type of discourse community. 3. Contrast and set up the macro and micro linguistic features they share, i.e., beyond rhetorical structures. The best method to test our hypothesis seems to be the Corpus Linguistics approach since we need to look for regularities. First, we will describe the genre of TNs. Second, we will focus on their register, namely their situational context, linguistic features, and functional relationships between such aspects (Biber & Conrad, 2019: 6) as they are found in a monolingual English corpus. Finally, we will contrast the results of the analysis of each corpus of TNs to highlight their similarities and differences. This twopronged approach will provide a comprehensive view of these specialized languages. Specifically, it will shed light on how communication is established in these specialized contexts and how the relationships between them are evidenced by them sharing features of genre and register. 2. The genre perspective The term ‘genre’ has been defined in a variety of ways, always connoting sameness in kind, type or form and function (Swales, 1990, 2002, among others). This term has been broadly used to refer to “language use in a conventionalized communicative setting in order to give expression to a specific set of communicative goals of a disciplinary or social institution, which give rise to stable structural forms by imposing constraints on the use of lexico-grammatical as well as discoursal resources” (Bhatia, 2004: 23). Analysing genre implies focusing “on the linguistic characteristics that are used to structure complete texts” which occurs mainly once in a text and serves a crucial role in how texts from a particular variety are constructed (Biber & Conrad, 2019: 15-16). Several scholars (Bhatia, 1993; Da Cunha & Montané, 2019; Parodi, 2010) Ibérica 43 (2022): 205-234 207 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO have noted that, even though certain genres can be viewed as belonging to specific specialized languages, such as recipes for the culinary domain or laws for the legal domain, other genres—including research papers, abstracts, reports, formal letters, or theses—are crosscutting, extending beyond one specialized context and remaining largely unchanged across disciplines. Accordingly, when a genre appears in more than one specialized context it is expected that there will be similar rhetorical structures, such as: the rhetorical organization of the genre, and the frequency of occurrence of moves and steps. However, as Parodi (2013) notes, few studies have undertaken such an analysis. In his article, he seeks to compare the frequency of occurrence of macro-moves, moves and steps in a corpus of university textbooks from four disciplines. The main findings of his study show that there are differences in the occurrence of some discourse moves and steps across the selected texts. He therefore concludes that disciplinarity plays a major role in the knowledge construction process and, consequently, in the way in which organizational discourse patterns are detected. 2.1. The Genre of Tasting Notes Professional discourses have their own specific characteristics that constrain their use and interpretation (Bhatia, 2008: 163). In this sense, TNs are considered to be professional genres that reproduce the tasting event in a textual form; writers need to describe what a wine or olive oil looks, smells, tastes, and feels like as faithfully as possible, while, simultaneously, writing in a way that may be understood by the genre’s growing readership (Caballero, 2017: 69). In this sense, TNs2 are typically organized into three different sections (moves, according to ESP approaches) which correspond to the three steps in any tasting procedure: “the assessment of wine’s colour, its smell (metonymically referred to as its ‘nose’), and its mouth-feel (a stage that involves smell, taste, and touch, which is metonymically referred to as the wine’s ‘palate’, and may be ‘de-composed’ into several stages)” (Caballero, 2017: 69). 3. The register perspective Register, characterized as the “expression-plane” of genre (Martin, 1985, as cited in Heid, 1994: 501), is more concerned with typical linguistic choices within different genres. Furthermore, it functionally connects those features 208 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE to the situational context of the variety; the choices are then seen as resulting from the contextual variables of topics/actions of language, participants/relationships, and textual organization (Biber & Conrad, 2019: 22). Because the focus is on words and grammatical features that are frequent, register analysis is pervasive. A situation analysis of wine and olive oil tasting notes3 points to three discourse communities among TNs writers: the professional taster, the amateur4, and the journalist and/or critic writing for the readers of specialized magazines (Peynaud, 1987). Peynaud discusses several ways of talking about the taste of wine depending on the circumstances, training and the taster’s state of mind (1987: 163): • Experts seek to express themselves with clarity and precision above all else. Their style is strict and economical but the comments they make are reasoned. The experts’ conciseness is not due to a lack of imagination, but stems from choosing the most precise words, and in their reports they only use those terms with an accepted and agreedupon meaning within their specialized context. In spite of their skills, their language should be simple and intelligible to all. • The two other discourse communities, the critic, “[t]he more occasional taster”, and “the informed amateur” do not always express themselves precisely. Their vocabulary is more limited, their style full of imagery but has less accuracy. They speak in metaphors and allusions, and not always in good taste. The inventiveness of their vocabulary conceals its vagueness. A comparison of the three different types of TNs reveals differences (see López-Arroyo & Roberts, 2016) in the knowledge the participants share on the topic, the social role they play, and the communicative goal. Therefore, it seems obvious that such a variety of writers would produce different types of TNs, especially in terms of their micro and macro linguistic features. In this sense, and as Ishizaki and Kaufer (2012: 276) state, “the micro (surface) linguistic choices of a text contribute to the textual overall features”; hence, if the different type of TNs writers use different language resources where different communicative goals may be assumed, and differences in the overall organization of the genre may be found out. In this sense, López-Arroyo & Roberts (2016: 373) argue that WTNs written by authors with different profiles show macro and micro linguistic Ibérica 43 (2022): 205-234 209 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO differences. However, all three groups face the same problem: to understand the meaning of and express, in words, the subjective sensations of smell and flavour evoked when tasting wine or olive oil. In a previous study, Sanz-Valdivieso & López Arroyo (2020: 31) found the same three profiles in OTNs. That means that specialized languages of wine and olive oil tasting notes share the same genre, regardless of author-type. However, to what extent do they share register features? To answer this, this paper will define the linguistic features of the TNs in these two specialized languages in order to be able to compare them and prove our hypothesis. 3.1. Register analysis of TNs The linguistic analysis of a register is based on register features from all linguistic levels: words, grammatical characteristics or syntactic constructions that are: “(1) pervasive (distributed throughout a text from the register), and (2) frequent (occurring more commonly in the target register than in most comparison registers)” (Biber & Conrad, 2019: 54) in a text variety and that are associated functionally with a given situational context in order to identify the language features that are typical or characteristic of the target register. A basic concern, therefore, is how to determine whether a linguistic feature is “typical” in a given register. Biber and Conrad (2019: 52) describe what they call “three major methodological considerations” to determine typicality: (1) the need for a comparative approach; (2) the need for quantitative analysis; and (3) the need for a representative sample of texts. One approach to studying register is to focus on a particular aspect of language use and compare it across registers. 4. Genre and register analysis in wine and oil tasting notes In this paper, we will compare and analyse one of the essential keys to the study of specialized languages: Phraseological units (Pus). Pus are word combinations or multiword units—lexical collocations involving verbs, nouns, adjectives such as verb + noun, adjective + noun, noun + verb, etc.—used in specialized discourse (L’Homme, 1995: 143). Since it is difficult to agree on a suitable definition of these units, as they cover different realities for different linguists, we will not be engaging with this because it is beyond the 210 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE scope of this paper. We understand that Pus is an umbrella term that covers a range of subtypes ranging from idioms and compounds that present a high degree of semantic and syntactic fixedness and are institutionalized to collocations and other units identified as lexical bundles, colligations, etc., which present some degree of lexical restriction but little fixedness and are not usually institutionalized (López Arroyo & Moreno-Pérez, 2019: 36). The flexibility of this concept provides ample opportunity for the study of phraseology of the wine and olive oil tasting specialized languages. By analysing Pus in wine and oil TNs, we will cover different register features varying from lexis to grammar and syntax. López-Arroyo & Moreno-Pérez (2019) describe the form and function of Pus, which they call lexical chunks, in WTNs and distinguish between five different categories based on recurrent patterns on the grammatical, textual, functional and collocational level: text organizing patterns, grammatical patterns, term forming patterns, term embedded collocations, and lexical collocations. It can be deduced that in that previous paper, it was not distinguished between genre and register features nor were they associated with the functional or situational context in which the TNs were produced and one of the purposes of the present paper is to cover that gap. In this paper we will examine some genre features, as repetitive structures prescribed in writing TNs—text organizing patterns. Additionally, we will also describe register features at a: 1. grammatical and syntactical level: recurrent grammatical structures, e.g., passive voice. 2. Lexical level: term-embedding collocations, which are verb-based structures that denote “what one can typically do with (or to) the object denoted by the base noun” (Martin, 1985 as cited in Heid, 1994: 238), such as to coat the mouth or to linger on the finish; term-forming patterns or multi-word terms, which are described as “collocates of a generic term” that add a higher degree of specificity, such as full body or zippy acidity; and lexical collocations, which are routine formulae that are not constructed around terms identified thorough recurrence, such as in the aftertaste or with hints of. Ibérica 43 (2022): 205-234 211 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO 5. Materials 5.1. Corpus and corpus design As Biber and Conrad (2019) suggest (see Section 3.1 above), the best way to describe linguistic features is through a representative sample of texts; in other words, a corpus. Corpora are known to be valuable resources to carry out quantitative and qualitative analyses of real utterances of a language in the context in which they are produced. Since there were no pre-existing corpora for this study’s chosen genre and specialized languages, we built an ad hoc domain-specific monolingual corpus (Corpas & Seghiri, 2009: 78). That is to say, a representative, reliable compilation (Seghiri, 2015: 142) of WTNs and OTNs originally written in English. Another criterion considered when designing this corpus was to include samples from the different communicative situations in which TNs are produced, that is to say, by the different types of writers described above (see Section 3). Texts written by amateur tasters were taken from blogs in which enthusiasts, newbies, influencers, and general consumers exchange their opinions about wine or olive oil. Specifically, we took these texts from, for instance, Olive Oil Online Forum, Wine Spectator, and Wine Follies; the writers of the samples in this sub-corpus will be referred to as Bloggers. Texts written by wineries or olive oil press companies were taken from the websites registered to official and institutional wine and olive oil webpages such as the California Olive Oil Council or Vintners Quality Alliance (VQA) of Ontario, just to name a few examples. This type of writer will be referred to as “Producers”. Finally, tasting notes written by Critics were taken from international contests or reputable critics like the Wine Advocate, published by Robert Parker. Our sub-corpora (see Table 1 below) consist of 620 olive oil and 251 wine TNs. The difference in the number of samples is due to their length, wherein olive oil TNs are much shorter than wine TNs. Our aim was to design a balanced corpus regarding both the number of tokens (21,105 and 19,899 respectively) and, particularly, the number of samples by each type of writer. 212 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE Corpora size OTNs corpus WTNs corpus Type of writer Samples Types Tokens Type of writer Samples Types Tokens Producers 250 1,223 8,933 Producers 90 1,414 7,056 Critics 230 720 9,779 Critics 90 1,754 7,487 Bloggers 140 559 2,322 Bloggers 71 1,282 5,356 TOTAL 620 1,726 21,105 TOTAL 251 3,097 19,899 Table 1: Size of OTN and WTN corpora. It was important to assess whether the a priori design criteria had been effective as to achieve representativeness—a notion that still remains controversial among corpus linguistics experts (Flowerdew 2004: 18). To do this, two tests were performed: the first consisted of two statistical and sampling formulae, allowing us to verify the quantitative representativeness of our corpus according to the quantity of samples and tokens in each one—the Confidence Interval (CI) of the Mean and the Standard Error of the Mean (SEx̄) - Our OTNs sub-corpus’ CI was calculated through the formula x±z s/√n, where x̄ is the mean (33.47191011236 in our corpus); Z is a value from 80% to 99.9%; chosen to calculate the desired value; s is the standard deviation according to the variable of text-length (18.391563599078 in our corpus); and n is the number of observations (620 samples). The resulting number was 33.47±2.43 for a 99.9% CI, from where it can be calculated that the Standard Error of the Mean (SEx̄) equals to 0.737 in our OTNs sub-corpus. - In the case of the WTNs sub-corpus, x̄=80.776892430279, z is chosen to be 99.9%, s=35.178033258049 and n=251, in which case CI=80.776892430279 ±7.306 and SEx̄= 2.220. These results can be interpreted in sample sizing as showing the corpus to be highly representative in quantitative terms. In fact, Biber (1993: 248) states that “the smaller this interval is the more confidence a researcher can have that she is accurately representing the population mean.” Secondly, in order to ensure the quantitative representativeness of our Ibérica 43 (2022): 205-234 213 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO corpus we used Recor, a software developed by Corpas et al. (2007) which performs an N-Cor algorithm-based quantitative analysis, using the type/token ratio to determine the representativeness of the corpus in terms of its size and lexical density (Seghiri, 2016: 386). The application generated two graphs5: • graph A shows the number of files on the horizontal axis, and the types/token ratio on the vertical one. It shows how many texts are needed for a corpus to be representative. • graph B displays the number of tokens on the horizontal axis. This graph can be used to determine the total number of tokens that are needed for the minimum size of a collection. Both graphs show an exponential decrease as the number of texts selected increases (Seghiri, 2015: 127). The corpus is determined to be representative at the point where the red and blue lines stabilize. It is here that we can observe the number of texts to include (Corpas & Seghiri, 2009: 127-128). According to our analysis using Recor, our corpus starts to be representative when there are 390 OTNs notes totalling approximately 12,500 tokens, and at 130 WTNs totalling 11,300 tokens. 5.2. Corpus Annotation Figure 1: OTNs representativeness graph output by Recor. 214 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE Figure 2: WTNs representativeness graph output by Recor Tagging and labelling a corpus implies adding its respective linguistic and extralinguistic information so that queries and data retrieval can be carried out quickly and accurately, otherwise the process “would be almost unimaginable” (McEnery & Hardie, 2012: 27-31). When labelling the samples, we followed the basic criteria by adding extralinguistic information so that they could be easily identified (see example 1 below). (1) “0108_OT_PR_OPR_Mz_18_EN” where 108 is the ID number; OT is the genre, in this case olive oil tasting note; PR the (Producers) sub-corpus; OPR the writer (California-based mill The Olive Oil Press); Mz the type of olive variety (Manzanilla); 18 the year of publication (2018); and EN (English) the language. We also tagged our samples grammatically, using a part of speech (POS) tagger designed for our purposes by the ACTRES (Análisis Contrastivo y Traducción English-Spanish/Contrastive Analysis and Translation EnglishSpanish) team. 6. Methodology and results 6.1. Genre perspective: Methodology and results All the texts contained in the corpus were tagged using pertinent rhetorical labels (moves and steps, according to Swales, 1990, 2002) to allow for a more in-depth analysis. According to Swales, a move is “a discoursal or rhetorical Ibérica 43 (2022): 205-234 215 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO unit that performs a coherent communicative function in a written or spoken discourse” (2002: 228-229); a step is a division of a move. Move analysis is a cognitive task which cannot be automatized. First, the “big picture” needs to be established by determining the overarching rhetorical purpose of the work. Second, the functionality of each section of the sample should be examined to determine its local purpose (Biber et al., 2007: 33). To describe this big picture, we identified the basic functional-semantic purposes of the tasting notes and grouped them together to reflect the moves and steps within each move (2007: 33). We took the three main sections—colour, smell, and mouthfeel (see Section 2.1 above)—of the tasting process and their function in the text as a starting point of our analysis (Caballero, 2017: 69). At this point, we referred to previous rhetorical studies on OTNs (SanzValdivieso & López-Arroyo, 2020) and WTNs (López-Arroyo & Roberts, 2014), where a rhetorical structure was identified for the tasting notes in both contexts. These structures were checked against our corpora and helped us narrow down the information sections we intended to identify; thereby, making it easier to create additional tags for steps of each move. This analysis was performed for both TNs and the different types of writers (see Table 2 below). However, some moves and steps occurred more frequently than others (López-Arroyo & Roberts, 2014) and, consequently, we decided to apply Suter’s distinction (1993) of obligatory and optional information. Suter (1993: 119) divides these types of information, based on the frequency of semantic units, into five categories: obligatory information (80-100%); high priority optional information (60-80%); medium priority optional information (40-60%); low priority optional information (20-40%); and occasional information (20% or less). Below, Table 2 shows the rhetorical structure. The moves are marked one, two, three, etc., the steps are identified as a, b, c, etc., and compulsory and high priority moves are in bold. Steps which are compulsory and high priority, given that the move they belong to occurs in a TN, are underlined. Percentages were calculated after quantifying all items referring to each move and step, as in the following rhetorically tagged OTN: [It is a strong and robust oil <Intensity>] [that is characterized by a soft green entry <Entry>], [aromas of freshly cut grass <Aroma>], [slight bitterness <Bitterness>] [and is mildly pungent <Pungency>]. The tags from all the subcorpora were counted, aggregated, and normalized on a base of 100 to know the percentage of each move and step present in our samples. 216 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE OTNs WTNs Prod. Critics Blog. Prod. Critics Blog. 1. Colour 10.80% 13.47% 11.43% 24.44% 46.67% 54.90% a. Hue, depth 92.59% 96.77% 93.75% 100% 100% 100% b. Clarity 3.70% 64.52% 25.00% 50.00% 43.90% 21.43% c. Density 14.81% 3.23% 25.00% 4.55% 19.51% 0% 2. Aroma 47.60% 79.13% 44.29% 71.00% 55.56% 73.24% a. Aroma 89.08% 93.96% 83.87% 100% 96.00% 100% b. Intensity 17.65% 20.33% 29.03% 28.13% 28.00% 9.62% c. Maturity 17.65% 15.38% 6.45% 34.38% 0% 0% 3. Taste 99.20% 99.57% 100% 100% 100% 98.59% a. Flavours 92.34% 94.76% 78.57% 98.89% 88.89% 95.71% b. Finish 41.53% 32.75% 22.86% 61.11% 32.22% 48.57% c. Pungency 56.45% 80.35% 34.29% 38.89% 45.56% 25.71% d. Mouthfeel 29.03% 31.44% 30.00% 81.11% 78.89% 60.00% e. Balance 32.26% 38.86% 15.00% 28.89% 13.33% 8.57% Table 2: WTN and OTN rhetorical structure and distribution. 6.2. Register perspective: Methodology and results The rhetorical labelling of the texts also allowed us to identify keywords that are typical of the different moves and steps in TNs as well as find the Pus that are pervasive and frequent in the tasting register. In order to do this, we followed a method developed by Thomas (1993: 47): “[O]ne way to determine the keyword or headword is […] to find which word takes ‘precedence’”. We looked for keywords, preferably nouns or verbs as headwords, because Pus tend to occur under a headword previously defined as a term in a specialized subject field (L’Homme, 1995: 239). We identified the candidate terms found in the corpus using the automatic term extractor TermoStat (Drouin, 2010) which compares specialized corpora to a reference corpus of general language: half from the British National Corpus and half from newspaper articles from the Montréal daily newspaper La Ibérica 43 (2022): 205-234 217 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Gazzette. The term extraction process starts by automatically adding POS tags to the corpus, upon which a filter is applied to “match different predefined syntactic matrices” such as noun + noun or adjective + noun. At the end of the process, the software allows for the retrieval of term candidates including data about their frequency, specificity, POS, concordance lines, etc. (Drouin, 2010). The automatic output was examined manually to remove the noise6, allowing us to obtain a wordlist of terms found in the whole corpus with, at least, five occurrences in each corpus: 69 candidates were found in the OTNs out of which 54 are nouns and 15 are verbs and 91 candidates were found in the WTNs, of which 58 are nouns and 33 are verbs. The next step consisted of examining our corpora with the aim of identifying the words, mainly nouns and adjectives, that collocate with these term candidates. To carry out this step we used Lancsbox, a tool for monolingual corpus analysis focusing on concordancing and text analysis. This allowed us not only to identify lexical items that frequently collocate with a given noun or verb, but also to know how frequently that association happened, and at which point in the text it occurs. After these steps, 129 different Pus were identified in the case of OTNs and 149 for WTNs. See appendix 2 for a list of all Pus common to both the OTNs and WTNs subcorpora. Below we include some examples of the Pus found in both corpora (see Section 4 above for a definition of the types of Pus): - Term-embedding collocations: to display fruitiness and to open the nose (OTNs critics); to open with aromas (WTNs producers). - Lexical collocations: in the aftertaste (OTNs critics); with hints of (OTNs producers, critics and bloggers, and WTNs producers); on the nose (OTNs and WTNs producers, critics and bloggers); a blend of (OTNs producers and critics, and WTNs critics). - Term-forming patterns or multi-word terms seem to be the most prolific category in both corpora, with instances such as green fruit (OTNs producers); flavor profile (OTNs and WTNs producers); bottle bouquet, lively entry and supple tannin (WTNs producers). Regarding grammatical features, we also found several recurring grammatical patterns in OTNs, which vary from irregular sentence structures to passive voice, the imperative or the simple present tense: 218 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE - Verbless phrases: Delicate aromas and flavors of butter and pure olive fruit, with a zesty pepper finish (OTNs bloggers). - Passive voice: It is produced in late October; harvesting is performed by hand (OTNs producers). - Present simple verbal tense: It is spicy with hints of artichoke and has a medium-intense aroma with grassy tones; Its aroma is clean and complex (OTNs producers). In WTNs, the recurrent grammatical patterns are: - Passive voice: The entry is honeyed with ripe citrus (WTNs producers). - Present simple tense of verbs: The wine is chewy on entry (WTNs producers). - Imperative mode: Savor and enjoy! (WTNs producers). 7. Discussion 7.1. Genre perspective: Discussion of results Once the quantitative analysis was carried out, we started the functional interpretation of the results “from description to an account of why these patterns exist” (Biber & Conrad, 2019: 69) in order to determine linguistic similarities and differences in the genre and register in both TNs and across the three different types of writers. Table 3 shows the rhetorical choices reflected in our corpora, for both OTNs and WTNs, and according to each type of writer: Ibérica 43 (2022): 205-234 219 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Olive oil tasting notes Wine tasting notes Producers Critics Bloggers Producers Critics Bloggers Colour 10.80% 13.47% 11.43% 24.44% 46.67% 54.90% Aroma 47.60% 79.13% 44.29% 71.00% 55.56% 73.24% Taste 99.20% 99.57% 100% 100% 100% 98.59% A+C 7.20% 11.74% 6.43% 23.33% 31.11% 36.62% A+T 47.20% 78.70% 44.29% 71.11% 55.56% 71.83% C+T 9.60% 13.04% 11.43% 24.44% 46.67% 38.03% Preferred T or A + T A+T T or A + T A+T A + C + T or T A + T or A+C+T Table 3: OTN and WTN move inclusion and preferred rhetorical combination. TNs written by olive oil and wine producers have aroma and taste as compulsory moves and their preferred structures include either only taste, or both aroma and taste. Perhaps this is due to their marketing strategy, which targets a particular audience so that the lay consumer understands the description of their products and is more likely to buy them. Amateur writers’ (bloggers) preferred rhetorical structure differs between OTNs and WTNs. In the case of olive oil, only taste or taste and aroma are included, while WTNs combine at least aroma and taste, with the possibility of including all three main moves. Olive oil critics include taste and aroma the most in OTNs, maybe because of their higher expertise on the subject. In the case of WTNs, critics include taste as their only compulsory move, although they also mention colour and aroma to a fairly significant degree. In any case, the three writer profiles of both specialized languages seem to share the view that taste prevails over colour and aroma. Colour is discussed the least in all cases. Besides, taste and aroma are the most strongly associated moves in both OTNs and WTNs, their combination being the most frequent in any writer profile. We then classified and contrasted the text-organizing patterns identified in each corpus, according to their form and function. These Pus are the most surface-level manifestation of textual organization reflected above, but the different information units included in TNs appear to be implicitly arranged 220 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE according to each main move (colour, aroma, taste) by means of other types of Pus and are not explicitly signalled by text-organizing patterns. Both corpora show a similar usage of text-organizing patterns. Producers use this resource the most in their texts, sharing the metonymical expressions (nose, palate) which also match the identified rhetorical structure (see Table 2 above). There is one exception to these matches: tasting note appears a total of 46 times in the WTNs corpus; however, all of these occur in TNs written by Robert Parker (critics sub-corpus). Thus, we may assume this cannot be taken as a feature of the genre, but rather as part of the individual style of that author. 7.2. Register perspective: Discussion of results Next, we studied the Pus according to the type of writer, their phrasal category (nominal phrase, prepositional phrase, verbal phrase) and the POS (Part of Speech) combinatorial pattern (adjective + noun, noun + preposition, etc.). The results enabled us to identify the function of those linguistic choices in relation to the situational context (see Table 4 below): Ibérica 43 (2022): 205-234 221 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO PU type Writer profile Term-embedding collocations Lexical collocations Term-forming patterns Total Type7 9/21 42.86% 25/36 69.44% 53/72 73.61% 87 Token 32/84 38.10% 214/649 32.97% 247/836 29.55% 493 Type 14/21 66.67% 33/36 91.67% 53/72 73.61% 100 Token 51/84 60.71% 376/649 57.94% 523/836 62.56% 950 Type 1/21 4.76% 21/36 58.33% 31/72 43.06% 53 Token 1/84 1.19% 59/649 9.09% 66/836 7.89% 126 Type 21/129 16.28% 36/129 27.91% 72/129 55.81% 129 Token 84/1,569 5.35% 649/1,569 41.36% 836/1,569 53.28% 1,569 Type 11/20 55.00% 19/23 82.61% 81/106 76.42% 111 Token 34/53 64.15% 106/237 44.73% 238/487 48.87% 378 Type 4/20 20.00% 16/23 69.57% 50/106 47.17% 70 Token 6/53 11.32% 66/237 27.85% 170/487 34.91% 242 Type 9/20 45.00% 17/23 73.91% 35/106 33.02% 61 Token 13/53 24.53% 65/237 27.43% 79/487 16.22% 157 Type 20/149 13.42% 23/149 15.44% 106/149 71.14% 149 Token 53/777 6.82% 237/777 30.50% 487/777 62.68% 777 Producers Olive oil Critics Bloggers Total Producers Wine Critics Bloggers Total Table 4: OTN and WTN distribution of PUs per class and writer profile. The results show that the type/token ratio is different among classes of Pus but similar in both specialized languages. Term-forming patterns seem to be the class of Pu most variedly and profusely used in both OTNs and WTNs, followed by lexical collocations, and lastly term-embedding collocations. OTN and WTN writers show the same tendency: all three writer profiles use term-forming patterns and lexical collocations rather than term-embedding collocations. The reason could be that TNs are by design more contentoriented than action-oriented texts. In our corpus, verb-based Pus (see OTN A. below) do not seem to be the most useful resource for writers to describe and/or evaluate a product. Instead, term-forming patterns (see OTN B. below) are well suited to this purpose, given that they are collocates of generic terms. The purpose of these terms is usually to add specificity, which is a function that answers the communicative needs of the tasting 222 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE discourse community. As can be seen by comparing OTNs A. and B. below, the same sensory expressions related to the finish of the olive oil can be achieved by omitting the verb: A. A touch of ripe tomato and celery with a lingering pungent and bitter finish and hints of bitter almond. B. With a subtle nutty taste, this buttery oil finishes with a pleasant bitterness. Likewise, lexical collocations, as recurrent routine formulae, serve the purpose of connecting content across the text, given that they are not tied to any specific specialized context or rhetorical section of TNs, as seen in WTN C. below: C. Delicate and juicy sweet dessert wine with a hint of sweet pears and mangos on the finish. These results match those found after we manually analysed all Pus according to the phrasal category they belong to and to their POS combinatorial pattern. Additionally, the phrasal categories of these Pus coincide for both TNs: the most common was the nominal phrase (NP), followed by the prepositional (PP), and the verbal phrases (VP) as shown in Table 5 below. Ibérica 43 (2022): 205-234 223 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Phrasal category Writer profile NP PP VP Total Type 64/86 74.41% 14/22 63.64% 9/21 42.86% 87 Token 365/1,206 30.27% 96/279 34.41% 32/84 38.10% 493 Type 66/86 76.74% 20/22 90.91% 14/21 66.67% 100 Token 752/1,206 62.35% 147/279 52.69% 51/84 60.71% 950 Type 39/86 45.35% 13/22 59.09% 1/21 4.76% 53 Token 89/1,206 7.38% 36/279 12.90% 1/84 1.19% 126 Type 86/129 66.67% 22/129 17.05% 21/129 16.28% 129 Token 1,206/1,569 76.68% 279/1,569 17.78% 84/1,569 5.35% 1,569 Type 89/116 76.72% 11/13 84.62% 11/20 55.00% 111 Token 304/627 48.48% 40/97 41.24% 34/53 64.15% 378 Type 59/116 50.86% 7/13 53.85% 4/20 20.00% 70 Token 218/627 34.77% 18/97 18.56% 6/53 11.32% 242 Type 43/116 37.07% 9/13 69.23% 9/20 45.00% 61 Token 105/627 16.75% 39/97 40.21% 13/53 24.53% 157 Type 116/149 77.85% 13/149 8.72% 20/149 13.42% 149 Token 627/777 80.69% 97/777 12.48% 53/777 6.82% 777 Producers Olive oil Critics Bloggers Total Producers Wine Critics Bloggers Total Table 5: OTN and WTN distribution of PUs per phrasal category and writer profile. Bloggers, in both specialized languages, do not use NPs as frequently as the other type of writers. Nevertheless, olive oil critics and wine producers use them more frequently than wine critics and olive oil producers. Regarding PPs, WTNs show a slightly smaller variety of Pus than OTNs. While olive oil producers use fewer PPs than their critics, wine critics utilize a less varied pool of PPs than wine producers. Both olive oil and wine bloggers’ preferred phrasal category appears to be PPs, both in variety and frequency of use, although wine amateurs use these phrases more than olive oil amateurs. VPs are the least variedly employed category across the writers in both corpora; producers again use this category more than the other two profiles. On the other hand, regarding the POS combinatorial patterns, Table 6 below shows those found in at least 5 different Pus and whose frequency accounts for at least 10% of the patterns used by each writer profile in our corpora. 224 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE Writer profile POS pattern Total Producers Critics Bloggers Type Token Type Token Type Token Type Token A+N 41/87 47.13% 184/493 37.32% 41/100 41.00% 342/950 36.00% 26/53 49.06% 51/126 40.48% 57/129 44.19% 577/1,569 36.78% N+N 9/87 10.34% 46/493 9.33% 8/100 8.00% 109/950 11.47% 2/53 3.77% 7/126 5.56% 10/129 7.75% 162/1,569 10.33% V+P+N 3/87 3.45% 3/493 0.61% 7/100 7.00% 13/950 1.37% 0/53 0% 0/126 0% 10/129 7.75% 16/1,569 1.02% N+P 8/87 9.20% 99/493 20.08% 8/100 8.00% 207/950 21.79% 5/53 9.43% 19/126 15.08% 8/129 6.20% 325/1,569 20.71% P+D+N 4/87 4.60% 24/493 4.87% 6/100 6.00% 65/950 6.84% 4/53 7.55% 12/126 9.52% 6/129 4.65% 101/1,569 6.44% P+N 2/87 2.30% 5/493 1.01% 4/100 4.00% 8/950 0.84% 6/53 11.32% 11/126 8.73% 6/129 4.65% 24/1,569 1.53% P+N+P 5/87 5.75% 53/493 10.75% 5/100 5.00% 48/950 5.05% 2/53 3.77% 11/126 8.73% 5/129 3.88% 112/1,569 7.14% Producers Critics Bloggers A+N 62/111 55.86% 176/378 46.56% 38/70 54.29% 142/242 58.68% 28/61 45.90% 66/157 42.04% 82/149 55.03% 384/777 49.42% N+N 19/111 17.12% 57/378 15.08% 12/70 17.14% 28/242 11.57% 6/61 9.84% 11/157 7.01% 22/149 14.77% 96/777 12.36% V+P+N 5/111 4.50% 12/378 3.17% 3/70 4.29% 4/242 1.65% 6/61 9.84% 8/157 5.10% 13/149 8.72% 24/777 3.09% N+P 6/111 5.41% 54/378 15.08% 6/70 8.57% 28/242 11.57% 5/61 8.20% 22/157 14.01% 6/149 4.03% 104/777 13.38% P+D+N 5/111 4.50% 30/378 7.94% 4/70 5.71% 12/242 4.96% 5/61 8.20% 26/157 16.56% 6/149 4.03% 68/777 8.75% Table 6: OTN and WTN distribution of most relevant PUs per POS pattern and writer profile. There is a larger variety of patterns in OTNs than in WTNs; adjective + noun (A+N) is the most common combinatorial pattern, as well as the more widely used, followed by the pattern noun + preposition (N+P) in terms of frequency of use, and by noun + noun (N+N) in terms of number of Pus with that pattern. This logical preference for the structure A+N across different writer profiles and different types of Pus confirms the results found by Sanz-Valdivieso & López-Arroyo (2020) and by López-Arroyo & Ibérica 43 (2022): 205-234 225 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Roberts (2016: 12): “wine tasting notes […] are intended to describe wines and the obvious way to do so is by adding descriptors to the key words for different aspects of wine”. Finally, Table 7 below shows the recurrent grammatical patterns analysed in both corpora and in the three sub-registers: Writer profile Total Producers Critics Bloggers Wine Olive oil Wine Olive oil Wine Olive oil Wine Passive voice 62/250 0.248 47/90 0.522 42/230 0.183 48/90 0.533 15/140 0.107 40/71 0.563 119/620 0.192 135/251 0.538 Present tense 490/250 1.960 308/90 3.422 356/230 1.548 273/90 3.033 123/140 0.879 193/71 2.718 969/620 1.123 774/251 3.084 Imperative mode 3/250 0.012 18/90 0.200 0/230 0.000 8/90 0.089 0/140 0.000 2/71 0.028 3/620 0.005 28/251 0.112 TNs with verbless sentences 27/250 10.80% 9/90 10.00% 142/230 61.74% 8/90 8.89% 20/140 14.29% 20/71 28.17% 189/620 30.48% 37/251 14.74% TNs with all verbless sentences 17/250 6.80% 1/90 1.11% 22/230 9.57% 0/90 0% 57/140 47.41% 4/71 5.63% 96/620 15.48% 5/251 1.99% TNs with no verbless sentences 206/250 82.40% 80/90 88.89% 66/230 28.70% 82/90 91.11% 63/140 45.00% 47/71 66.20% 335/620 54.03% 209/251 83.27% Verbless sentences Olive oil Table 7: OTN and WTN distribution of most relevant grammatical patterns per writer profile. In the case of passive voice, present tense, and the imperative mode, we tried to obtain a mean per text; the numbers compared were calculated as follows: instances of passive voice/texts (i.e., 0.248 instances of the passive voice per text). In the case of verbless sentences, we examined TNs not only to calculate the mean for verbless sentences per text, but also to obtain information about whether each text was made up of any, some, or all verbless sentences, i.e., the number of TNs containing no verbless sentences/total number of tasting notes, for example. Producers use the passive voice and the present tenses more often than the critics, and critics employ these more regularly than amateur writers. This 226 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE seems logical, since the passive voice puts the focus on the object —olive oil, wine, and their qualities— and the present simple is the most obvious verbal tense for description and evaluation. However, these patterns can be found more recurrently, around twice as often, in WTNs than in OTNs. Results regarding the use of the imperative mode leave no doubt as to whether it is a characteristic grammatical pattern in WTNs, although not OTNs. Perhaps this is because wine requires more specific consumption practices and dates than olive oil in terms of temperature, pouring, conservation, etc. Only TNs written by olive oil critics include some verbless sentences—more than half of their texts. The second most common case is texts containing at least one verbless sentence. As for TNs made up entirely of verbless sentences, there is a considerable difference: WTNs rarely include verbless sentences; notably those written by wine critics do not include any at all. As noted, this is in sharp contrast to olive oil critics. However, the rest of the writer profiles show similar tendencies. This, coupled with the popularity of wine tasting in comparison to olive oil tasting, could be explained as OTNs writers seeking to make their texts accessible by employing such methods as grabbing the reader’s attention, emphasizing certain elements, and/or enlivening the text by using an informal or conversational style (LópezArroyo & Roberts. 2015: 166). In any case, these results are in line with those found in the analysis of Pus above: verbs appear in WTNs to a greater extent than in OTNs. 8. Conclusions This study of olive oil and wine tasting notes has proven that certain specialized discourses share crucial genre and register features: 1. Both have almost identical rhetorical structures in regards to moves and steps. Text-organizing patterns seem to be restricted to producers’ TNs, with writers generally organizing information implicitly, i.e., not using text-organizing patterns and simply sequencing information related to the different moves. Although there are some variations across the writer profiles, both olive oil and wine TNs share preferred rhetorical choices, meaning their texts show similar organizational patterns in their arrangement of the same information units. Ibérica 43 (2022): 205-234 227 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO 2. Both specialized languages share a register in which three distinct communicative goals can be identified. They are further characterized by a similar set of terms and Pus although they are employed to different extents. All writers preferred term-forming patterns and lexical collocations when expressing content-based descriptions and evaluations, mostly in the form of nominal phrases made up of A+N. Amateurs, however, seemed to prefer prepositional phrases and their overall use of Pus is scarcer and more irregular than that of producers and critics. 3. Producers and amateurs seem to behave similarly when writing TNs. This may be due to the need for producers to reach as many consumers as possible, regardless of their expertise or experience in the specialized context of tasting, thus bringing their register closer to that of their target audience. 4. There are more differences among the different writers of OTNs than among WTNs, which could point to the fact that the wine tasting discourse community is more homogeneous and less linguistically stratified than that of olive oil. Perhaps this is due to the fact that wine tastings have a long history and have attracted participants from across the socio-economic spectrum resulting in more laypeople having the ability to access the specialized language of wine tasting. This, however, does not seem to be the case with olive oil tasting, which may be a consequence of the newness of tasting olive oil as a hobby outside of specialized and professional contexts. In future studies, it would be interesting to focus on other linguistic features complementary to phraseology in an effort to explore whether or not there is a shared nature in the tasting register and sub-registers. In any case, TNs authors writing in English need to carefully consider their intended readership, not only because of the shared features of olive oil and wine tasting specialized languages, but also due to the different writer profiles, to ensure that their texts match the expectations of the target discourse community and are accepted as part of the genre of tasting notes. Article history: Received 9 February 2022 Received in revised form 13 April 2022 Accepted 23 April 2022 228 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE References Bhatia, V. K. (1993). Analysing genre: Language use in professional settings. Routledge. Bhatia, V. K. (2004). Worlds of written discourse: A genre-based view. Continuum. Biber, D., Connor, U., & Upton, T. A. (2007). Discourse on the move: Using corpus analysis to describe discourse structure. John Benjamins. Biber, D., & Conrad, S. (2019). Register, genre and style. Cambridge University Press. Caballero, R. (2007). Manner-of-motion verbs in wine description. Journal of Pragmatics, 39, 20952114. <https://psycnet.apa.org/doi/10.1016/j. pragma.2007.07.005> Caballero, R. (2017). From the glass through the nose and the mouth: Motion in the description of sensory data about wine in English and Spanish. Terminology, 23(1), 66-88. <https://doi.org/ 10.1075/term.23.1.03cab> Corpas Pastor, G., & Seghiri Dominguez, M. (2009). Virtual corpora as documentation resources: translating travel insurance documents. In A. Beeby, P. Rodríguez & P. Sánchez Gijón (Eds.), Corpus use and translating (pp. 75-107). John Benjamins. Corpas Pastor, G., Seghiri Dominguez, M., & Maggi, M. (2007). Recor. Software. http://www. lexytrad.es/es/recursos/recor-2/. Da Cunha Fanego, I. & Montané March, A. (2019). Textual genres and writing difficulties in specialized domains. Revista signos: estudios de lingüística, 55(99), 4-30. http://dx.doi.org/ 10.4067/S0718-09342019000100004. Diederich, C. (2015). Sensory adjectives in the discourse of food: A frame-semantic approach to language and perception. John Benjamins. <https://doi.org/10.1075/celcr.16> Drouin, P. (2010). TermoStat Users Guide. <http://termostat.ling.umontreal.ca/doc_termostat/ doc_termostat.html> Flowerdew, L. (2004). The argument for using English specialised corpora to understand academic and professional language. In U. Connor & T. A. Upton (Eds.) Discourse in the professions: Perspectives from corpus linguistics (pp. 11-13). John Benjamins. Heid, U. (1994). On ways words work together topics in lexical combinatorics. In W. Martin, W. Meijs, M. Moerland, E. ten Pas, P. van Sterkenburg & P. Vossen (Eds.), EURALEX’94 Proceedings (pp. 226-257). Vrije Universiteit Amsterdam. Isizhaki, S. & Kaufer, D. (2012). Computer aided rhetorical analysis. In P. McCarthy & D. Boonthum (Eds.). Applied natural language processing: Identification, investigation and resolution. Information science reference (pp. 275-295). Hershey. L’Homme, M. C. (1995). Processing word combinations in existing term banks. Terminology, 2(1), 141-162. <https://doi.org/10.1075/term. 2.1.08pro> López Arroyo, B., & Moreno Pérez, L. (2019). Lexical chunks in English and Spanish sales contracts: A corpus based study. Terminology, 25(1), 32-59. <https://doi.org/10.1075/term. 00027.lop> López Arroyo, B., & Roberts, R. P. (2014). English and Spanish descriptors in wine tasting terminology. Terminology, 20(1), 25-49. <https: //doi.org/10.1075/term.20.1.02lop> López Arroyo, B., & Roberts, R. P. (2015). Unusual sentence structure in wine tasting notes: A contrastive corpus based study. Languages in Contrast, 15(2), 162-181. <https://doi.org/ 10.1075/lic.15.2.01lop> López Arroyo, B., & Roberts, R. P. (2016). Differences in wine tasting notes in English and Spanish. Babel, 62(3), 370-401. <https://doi.org/ 10.1075/babel.62.3.02lop> McEnery, T., & Hardie, A. (2012). Corpus linguistics: method, theory & practice. Cambridge University Press. Montoro del Arco, E. T., & Roldán Vendrell, M. (2013a). Las denominaciones del aceite de oliva en los reglamentos oficiales: evolución, normalización y problemas terminológicos. In C. Sinner (Ed.), Comunicación y transmisión del saber entre lenguas y culturas (pp. 265-282). Peniope. Montoro del Arco, E. T., & Roldán Vendrell, M. (2013b). Terminología, normalización y comunicación: las categorías de aceite de oliva en español, inglés y chino. Terminology, 19(1), 62-92. <https://doi.org/10.1075/term.19.1.03mon> Montoro del Arco, E. T. (2012). La fraseología especializada del área científica y socioeconómica del olivar y el aceite de oliva: entre la tradición y la modernidad. In A. Pamies, J. M. Pazos Bretaña & L. Luque Nadal (Eds.), Phraseology and discourse: cross-cultural and corpus-based approaches (pp. 338-345). Schneider Verlag. Ibérica 43 (2022): 205-234 229 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Muegge, U. (2012). 10 Things you should know about automatic terminology extraction. The American Translators Association Chronicle, 41(9), 24-28. Paradis, C. (2010). Touchdowns in winespeak: Ontologies and construals in use and meaningmaking. In G. Rambaud & A. Poves Luelmo (Eds.), Proceedings from the 1st conference on linguistic approaches to food and wine descriptions (pp. 112). UNED. Parodi, G. (2010). The rhetorical organization of the textbook genre across disciplines: A ‘colony-inloops’? Discourse Studies, 12(2), 195-222. <https://doi.org/10.1177%2F1461445609356500> Parodi, G. (2013). Genre organization in specialized discourse: disciplinary variation across university textbooks. Discourse Studies, 16(1), 6587. <https://doi.org/10.1177%2F1461445613496 355> Peynaud, E. (1987). The taste of wine: Art and science of wine appreciation. The wine appreciation guild. Robinson, J. (2015). The Oxford companion to wine. Oxford University Press. Roldán Vendrell, M. (2007). Unidades de conocimiento especializado en textos del ámbito de la oleicultura antigua: aspectos sintagmáticos. In G. Águila Escobar, P. Barros García & E. T Montoro del Arco (Eds.), Estudios lingüísticos, literarios e históricos. Homenaje a Juan Martínez Marín (pp. 327-336). Universidad de Granada. specialised languages. Terminology, 18, 9-26. Santa María, S. (2013). The terminology of olive oil taste testing: a bilingual (Spanish-English and English-Spanish) glossary. Translation Journal, 17(2), 13-34. Sanz Valdivieso, L., & López Arroyo, B. (2020). On Describing Olive Oil Tasting Notes in English. Fachsprache. Journal of Professional and Scientific Communication 42(1-2), 27-45. <https://doi.org/10.24989/fs.v42i1-2.1825> Seghiri, M. (2015). Determinación de la representatividad cuantitativa de un corpus ad hoc bilingüe (inglés-español) de manuales de instrucciones generales de lectores electrónicos. In M. T. Sánchez Nieto (Ed.), Corpus-based translation and interpreting studies: From description to application (pp. 125-146). Frank & Timme. Seghiri, M. (2016). Compilación de un corpus trilingüe de seguros turísticos (español-inglésitaliano): Aspectos de evaluación, catalogación, diseño y representatividad. [Doctoral dissertation, Universidad de Málaga]. RIUMA. <https://riuma. uma.es/xmlui/handle/10630/2715>. Suárez Toste, E. (2017). On the roles of metaphors and synesthesia in wine reviews. Terminology, 23(1), 89-112. <https://doi.org/ 10.1075/TERM.23.1.04SUA> Suter, H. J. (1993). The wedding report: A prototypical approach to the study of traditional text types. John Benjamins. Roldán Vendrell, M. (2010). Olivaterm: Vocabulario multilingüe del aceite de oliva. In E. T. Montoro del Arco & J. A. Moya Corral (Eds.), El Español en contexto. Actas de las XV jornadas sobre lengua española y su enseñanza (pp. 312319). Universidad de Granada. Swales, J. (1990). Genre analysis: English in academic and research settings. Cambridge University Press. Roldán Vendrell, M. (2013). Diccionario de términos del aceite de oliva. Arco/Libros. Thomas, P. (1993). Choosing headwords from LSP collocations for entry into a terminology data bank. In H. B. Sonneveld & K. L. Loening (Eds.), Terminology. Applications in interdisciplinary communication (pp. 43-68). John Benjamins. Roldán Vendrell, M., & Fernández Domínguez, J. (2012). Emergent neologisms and lexical gaps in Swales, J. (2002). Research genres: Exploration and applications. Cambridge University Press. Belén López Arroyo is an Associate Professor in ESP at the university of Valladolid (Spain). She currently teaches ESP and translation and Corpus Linguistics in the English Studies Degree. Her research interests include Terminology, Contrastive analysis and Translation. She is author of several articles and books related to contrastive analysis of scientific and professional genres and its implication for translation. In the ACTRES team she is in charge of the Rhetoric of Expert-to-Expert Discourse (in different 230 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE areas) as well as of terminology and its applications for developing writing aids in English for Spaniards. Along with other members of the ACTRES group she is the author of semi-automatic writing generators aids in the field of Oenology. Lucía Sanz-Valdivieso has a Master’s in Education and in Advanced English Studies. She is currently a PhD candidate in the Advanced English Studies program and is a teaching assistant at the university of Valladolid (Spain). Her research is devoted to the study of sensory language and the development of digital assisted-writing tools for specialized and professional communication. She has published some articles in the fields of specialized discourse in English and English-Spanish contrastive analysis. NoTeS 1 WTNs and OTNs will be used to refer to wine and olive oil tasting notes respectively. See appendix 1 for an illustration of the content and organization of typical OTNs and WTNs, taken from our corpus. 2 For a complete description of the situational characteristics of register and genres, see Biber & Conrad, 2019: 40. 3 The term ‘amateur’ in wine tasting was introduced by Robinson (2015) to describe laypeople or beginners in the world of wine. 4 Recor software was developed by LexyTrad, a lexicography and translation research group of the university of Málaga, Spain. The software is developed in Spanish and therefore the legends in the graphs are available only in Spanish. 5 6 Invalid term candidates (Muegge 2012: 24). “Type” and “token” are used in tables 4, 5 and 6 only in relation to Pus, not to types and tokens in the whole corpus as described above. Like this, two occurrences of long finish are accounted as one type (one form) and two tokens (two instances of use) in an effort to take into account both variety and quantity. 7 Ibérica 43 (2022): 205-234 231 BELéN LóPEz ARROyO & LuCíA SANz-VALDIVIESO Appendix 1: examples of oTNs and WTNs by different writer profile Olive oil producer TN (sample 0211_OT_PR_NAP_AC_18_EN from our corpus): This oil is a classic, stone fruit, ripe Ascolano olive oil from California. It has a wonderful floral aroma and an intense peach-apricot-like taste with some nuttiness and butteriness too. It is very fresh, crisp, and zingy because of the green, herbaceous undertone qualities, which gives it a nice, teasing pungency. It is very fresh and the balance is perfect. Wine critic TN (sample 0005_WT_CR_HAB_RE_98_EN from our corpus): The dark garnet-colored 1961 Haut-Brion is pure perfection, with gloriously intense aromas of tobacco, cedar, chocolate, minerals, and sweet red and black fruits complemented by smoky wood. This has always been a prodigious effort (it was the debut vintage for Jean Delmas). It is extremely full-bodied, with layers of viscous, sweet fruit. This wine is akin to eating candy. Consistently an astonishing wine! Wine amateur TN (sample 0513_OT_BG_RUS_BL_18_EN from our corpus): Quite a rich, golden colour, the nose is subtle and refined with notes of melon, lemon, thyme and pepper. The palate is full and almost creamy with a tangy grip on the finish. 232 Ibérica 43 (2022): 205-234 TASTINg NOTES: A CORPuS-BASED STuDy OF OLIVE OIL AND WINE TASTINg DISCOuRSE Appendix 2: PUs present in both oTN and WTN corpora Text-organizing patterns Nose Term-embedding collocations Lexical collocations Term-forming patterns To add complexity Palate To open with aromas To open with scents Aroma of With a hint of On the palate With aromas of In the mouth In taste In color On the nose In texture On the finish In the nose With a touch of With hints of Notes of A touch of Hints of With notes of A hint of Nuances of Low acidity Floral notes Tropical fruit Floral aroma Clean finish Exotic fruit Fresh aroma Spicy finish Citrus fruit Intense aroma Long finish Red fruit Berry fruit Rich finish Stone fruit Full body Fruit flavor Velvety texture Aromatic nose Ripe flavor Rich texture Ibérica 43 (2022): 205-234 233