Lexical chunks in English and Spanish sales contracts: A corpus-based study

Lexical chunks in English and Spanish sales contracts A corpus-based study Belén López Arroyo & Leticia Moreno Pérez Universidad de Valladolid This paper focuses on the study of word combinations “of common usage” which are “lexicalized”, have “syntactic and semantic stability, may be idiomatized and carry connotations, and have an emphatic or intensifying function.” (Gläser 1994/1995,45). Following previous research on Languages for Specific Purposes (LSP) and legal phraseology, we will define, identify and classify these units in English and Spanish according to their form and meaning, using a comparable corpus of sales contracts. To carry out our study, we will focus on a number of descriptors that are commonly used within these units on the basis of the headwords they collocate with, in order to determine how specific or general they are in their form, use and meaning since this issue poses translation problems. As genres determine matters such as or terminology and phraseology, the results will be useful for specialized translators and legal drafters. Keywords: lexical chunks, sales contrasts, corpus-based studies 1. Introduction Thestudyofwordcombinationsormultiwordunits(i.e.lexicalcollocationsinvolvingverbs,nouns,adjectivessuchasverb+noun,adjective+noun,noun+verb, etc. (L’Homme1995,143)hasbecomeanessentialpartofthestudyofLanguageforSpecific Purposes (LSP), especially since corpus linguistics has started playing a key role inthestudyof language useandfunction.Indeed, “wordsdo notexistinisolation but cluster together in particular ways to make larger meaningful units of language” (Jones and Waller 2015,84) and such a selection does not occur randomly, but is formulaic in nature. In this sense, Sinclair states that about 80% of the words in discourse are used according to the co-selection principle rather than for purely syntactic or grammatical reasons (Sinclair 2000,197). Thus, the analysis of how https://doi.org/10.1075/term.00027.lop Terminology 25:1 (2019), pp. 32–59. issn 0929-9971 |e‑issn 1569-9994 © John Benjamins Publishing Company wordscombinewithotherwordstoformmultiwordunits(L’Homme1995)isanecessaryfocusofstudyforanylinguist. The interest in word combinations has also attracted the attention of terminologists who deal with specialized languages. Notes on phraseological units, for instance, are inserted in the context section of terminological records, which shows the term in its immediate environment and gives information on the typical usage, structure and meaning of the term as opposed to the meaning of their individual units (Cowie 1998,3168). Indeed, there is definitely a growing awareness of the fact that it is not enough to know the terminology of a discipline to write specialized texts in that discipline, but that one also needs to know the contextualized use of terms in order to produce technically appropriate texts either in the original language or in translation. In this sense, word combinations (multiword units) pose a problem for translators and English for Specific Purposes (ESP) writers since these units are linguistic and culturally bound (Temmerman 2017,133) and may vary between languages. The aim of this paper is to define, identify and classify word combinations in the language of real property sales contracts using an English-Spanish comparable corpus to provide assistance when writing or translating this genre. We will begin by defining the concept of “word combination” and presenting our approach to its study. In a particular LSP, most of its word combinations are terms;1hence, looking for keywords in a specialized field permits identifying and describing word combinations. As a consequence, in the present paper we will analyze the key sales contract terms looking for patterns regarding form, meaning and occurrence, and wewill classify themaccording totheirdegree ofsyntactic andsemantic fixedness. Finally, we will contrast the findings obtained in the two languages under study in order to offer translators and specialized writers of real property sales contracts guidelines to help them accomplish their tasks. 1. A distinction between words and terms seems to be necessary at this stage; word is a linguistic concept (…) defined as a syntagmatic unit which is located within the hierarchy of linguistics units. (…) Terms, however, are a functional class of lexical units. (…) What makes some lexical units terms is their usage and social recognition within a given domain, subject or vocation (Kageura 2017,46–47). Lexical chunks in English and Spanish sales contracts 33 © 2019. John Benjamins Publishing Company All rights reserved 2. Word combinations and approaches to their study The tendency of language components to cluster together in predictable ways has led some researchers to study word combinations in depth. However, even nowadays it is difficult to find a suitable definition of these units since they cover different realities for different linguists. Scholars agree that the phraseology of LSP has not been as studied as that of language for general purposes (LGP) (GoźdźRoszkowski and Pontrandolfo 2015; Dobrić Basaneže 2017). Its definition varies depending on whether LGP or LSP is being considered (Roberts 1998,62). Moreover “what is covered by the term word combination is to some extent language dependent” (Roberts 1998,63) since, for instance, some units are considered part of phraseology in one language (phrasal verbs in English, for example) and not in others (phrasal verbs do not exist in Spanish or French) (López Arroyo and Roberts 2016,3; Wright 1997,14). Also, “one unit in the group or the group as a whole conveys a meaning that cannot be easily understood, or usage in the target language is unknown” (L’Homme 1995,143). All these observations prove the complexity of this subject matter. What linguists agree upon is that the most prominent way to study word combinations is through collocations, “a combination of two lexical (as opposed to grammatical) words often found together or in proximity, e.g. make sense” (Timmis 2015,26). This means that when a word is used, there is a high statistical probability that a certain word (or words) will occur alongside with it. Some linguists name these patterns formulaic sequences (Schmitt and Carter 2004) while others have used terms such as chunks and lexical bundles (O’Keeffe et al. 2007; Biber et al. 1999; Timmis 2015; Jones and Waller 2015, among others) or phraseological units (Gläser 1998; Roberts 1998, among others). Traditional approaches have long dealt with multiword units, wherein a single meaning is attached to more than one word, for example, phrasal verbs, compounds, and idioms. But corpus-based research has shown that collocations extend far beyond the level of such units. “In fact, it appears quite common for longer sequences of words to pattern together. Some of these recur frequently enough to be treated as units in their own right, e.g. to make a long story short. Numerous terms have been coined to refer to this type of sequence, but the most commonly used are lexical chunks and lexical phrases.” (Schmitt 2000,400). In the present paper we will use the term lexical chunk to refer to these units, described by Timmis as “a frequent meaningful sequence of words that may include both lexis and grammatical words, e.g to a certain extent (includes a preposition and an article)” (2015,26–27). Gläser (1994/1995,45) completes this description stating that “they have a syntactic and semantic stability, may be idiomatized, may carry connotations, may have an emphatic or intensifying function in a text and are used in the discourse com34 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved munity of a particular subject field”. In addition, they may include phrasal verbs, idioms, compounds, collocations, etc., which have been traditionally recognized as word combinations or multiword units (Gläser 1998). Glaser (1994/1995,46–58) has studied lexical chunks (she calls them phraseological units or PUs) in English basing her classification on the notions of “centre” and “periphery” from the Prague school. The “centre” comprises phraseological units that function as single words such as nouns (blind alley), verbs (make assumptions, make choices), noun+adjective units (unconditional surrender), noun+prepositional phrase units (burden of proof), verb phrases (to give someone the benefit of the doubt), and function words that denote relations between phenomena or objects and functioning as adverbs (prepositions: by dint of, in terms of/ conjunctions: in order to, as if) and bilexemic units (blanket cover). The “transition area” includes ‘irreversible binomials’ (word pairs which have a fixed order such as goods and chattels), stereotyped combinations or similes (as cold as a cucumber), sentence-like fragments or sentence-like phrases (a rolling stone), quotations and literary allusions (Scylla and Charibdis situation) and finally, collocations (to produce an alibi / a certificate / a document). Last but not least, the “periphery” includes proverbs, truisms (boys will be boys), maxims, quotations, slogans and routine formulae (if death occurs) (Molina Plaza 2009,120). As a conclusion, we could state that lexical chunk is a generic term that covers a range of subtypes, which have been classified according to their degree of semantic fixedness, syntactic fixedness, lexical restrictions and institutionalization. These subtypes include, on the one hand, idioms and compounds that present a high degree of semantic and syntactic fixedness and are institutionalized, and, on the other hand, 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. 3. Lexical chunks in legal language When studying lexical chunks in LSP, Gläser (1994/1995) took into account different subject fields, including law, which will be the focus of this study and whose phraseology has been scarcely analyzed (Biel 2014; Pontrandolfo 2015). However, as mentioned above, this classification was developed for LSP in general. To focus on the field of law, we consider it necessary to adapt more specifically to the subject matter. Therefore, we will also use the classification developed by Biel (2014), which takes into account the nature of legal genres from the point of view of meaning – i.e. function. Biel proposed five categories: Lexical chunks in English and Spanish sales contracts 35 © 2019. John Benjamins Publishing Company All rights reserved –Text-organizing patterns: genre-dependent repetitive structures that build up the framework of the legal text and are sometimes prescribed for drafting, e.g. titles or closing formulas as whereas or acting in accordance with. –Grammatical patterns: genre-dependent repetitive grammatical structures, e.g. conditional clauses (in the event that) or passive voice (shall be deemed). –Term-forming patterns (or multi-word terms): structures comprised of terms that are specified by different collocates to create subtypes of such term; they usually follow the pattern Adj+Noun or Noun+Noun, and are transparent in meaning, e.g. European public limited-liability company. –Term-embedding collocations: verb-based structures, mostly Noun+Verb, that express actions related to terms, e. g. to hold shares. –Lexicalcollocations:importantrepetitiveformulaethatdonotrevolvearound termsandareidentifiedthroughrecurrence,e.g.subjecttothisRegulation. The combination of these two classifications will allow us to analyze our findings from a multi-faceted point of view, yet focusing on the characteristics of a specific genre, the real property sales contract, since terminology, syntax and discursive aspects strongly depend on text typology (Tabares Plasencia and Pérez Vigaray 2007,570). In other words, we will adopt Gläser’s (1994/1995) classification of lexical chunks to describe and classify their form and Biel’s (2014) typology to describe and classify them according to their meaning. 3.1 The genre of real property sales contracts Sales contracts are a specific genre in the field of law. A genre is defined as a class of communicative events which possess features of stability, name and recognition (Swales 1990,9). Exemplars of a genre exhibit patterns of similarity in terms of structure, style, content and intended audience. Therefore, the microand macrolinguistic levels (lexical chunks and rhetoric) are comprised in the notion of “genre” and are crucial concepts in our study. According to Borja’s classification of legal genres (2000,82), a sales contract or sale agreement2is a genre with an instructive function, which regulates an agreement between a buyer and a seller covering the sale and delivery of goods, securities, and other personal property. Alcaraz et al. (2007) state that it is one of the most important genres in the regulation of interpersonal activities and exchanges. Sales contracts are common in most legal systems. 2. In legal contexts, agreement and contract are used as synonyms although there are some differences between them. In Common Law, a contract is an enforceable agreement which includes an offer and acceptance and a consideration, whereas an agreement does not have these binding requirements. 36 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved Sales contracts must be in writing in order to be enforceable and should specify the parties involved in the subject matter to be sold, and any material or special terms or conditions. Some of them include the consideration, in other words, the amount and type of payment. Sales contracts have a fixed content in English and Spanish. What is more, while some legal systems do not specify their structure in terms of clauses and external structure, they define the contents to be included. All in all, sales contracts show an internal organization of information (Borja 2000,84). 4. Methodology 4.1 Corpus In this paper, we have adopted Sinclair’s bottom-up approach to lexical chunks (2000), i.e. using corpora as a tool “to identify specific discourse units within texts” (Biber et al. 2007,241) in order to classify findings through quantitative and qualitative analyses. The starting point of our study is an ad hoc domain-specific (Corpas and Seghiri 2009,78) comparable corpus, that is to say, a representative, reliable text compilation (Seghiri 2015,142) of real property sales contracts written originally in English and Spanish. We built a comparable corpus of similar real property sales contracts in English and Spanish by searching for real property sales contracts in the same type of source (López Arroyo and Roberts 2016,377). We started our search on the Internet in legal specialized websites, using the keywords real property sales contracts and contratos de compraventa. Then we restricted our search to law firm websites in the US and in Spain, law books, government websites and realtors (California Association of Realtors 2018; Florida Bar 2018, for example), obtaining original real property sales contracts as well as sample forms. The samples were all published between 2013 and 2015. This initial real property sales contracts corpus includes 58 samples in English and 181 in Spanish which amount to 204,685 and 265,800 words respectively. The size of the corpus is more than adequate for a specialized ad hoc corpus, according to Bowker and Pearson (2002,48) and Biber (1993,254). However, this corpus was still too large for this specific study, so we have used a more restricted corpus, consisting of a limited number of sample texts in English and Spanish. Given that the English real property sales contracts in our original corpus are longer than the Spanish ones, we included 10 sample texts in English and 18 in Spanish so as to balance the number of words in the two languages. The number of words in the samples we are using in the present study are 20,277 in English and 22,828 in Lexical chunks in English and Spanish sales contracts 37 © 2019. John Benjamins Publishing Company All rights reserved Spanish. The texts comprising our limited corpus were not chosen completely at random from the initial corpus, because we wanted to avoid including a number of texts from the same company or agency. Our selection process paid particular attention in ensuring that there was a balance regarding the writing styles used in real property sales contracts. In order to ensure the representativeness of our corpus we used Recor, a software developed by Corpas et al. (2007) for this purpose. The application generates two graphs: – Graph A shows the number of files on the horizontal axis, and the types/ tokens ratio on the vertical one; it is used to show how many texts should be set for the corpus to be representative. – Graph B shows the number of tokens on the horizontal axis. This graph can be used to determine the total number of words that should be set for the minimum size of the collection. Both operations show an exponential decrease as the number of texts selected increase. However, at the point where both the red and blue lines stabilize, it is possible to state that the corpus is representative, and at precisely this point it is possible to see approximately how many texts will produce this result (Corpas and Seghiri 2010,127–128). Graphs 1 and 2 show the corpus representativeness in English and Spanish. As we can see, 10 samples and 2,500 words are needed in English for the corpus to be representative and 2 samples and 5,000 words are needed in Spanish. According to these data, we can state that our corpus is representative for the analysis to be carried out. 4.2 Corpus annotation According to McEnery and Hardie (2012,27–31), corpus annotation and labelling, along with search tools, allow the researcher to explore linguistically motivated queries and retrieve linguistic data in a quick and accurate way “that would be almost unimaginable otherwise” (2012,28). In this particular case, the annotation of the sales contracts included in our corpus was an essential step to define the internal organization of the genre’s content. We annotated the texts using labels that indicate the rhetorical structure of sales contracts in general (moves and steps, according to Swales 1990,2002). All the texts contained in the corpus were annotated using pertinent rhetorical labels to allow for deeper analysis (López Arroyo et al. 2018). Table 1 shows the rhetorical structure. The moves are marked one, two, three, etc., the steps are identified as a,b,c, etc. 38 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved a. b. Graph 1. Representativeness in the English corpus Lexical chunks in English and Spanish sales contracts 39 © 2019. John Benjamins Publishing Company All rights reserved a. b. Graph 2. Representativeness in the Spanish corpus 40 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved Table 5. (continued) Key word Lexical chunks Occurrences Impuesto de Actos Jurídicos Documentados 3 Requerimiento Fehaciente 514. Requerimiento 37 Requerimiento Notarial o Judicial 3 15. Cargas 29 Cargas y Gravámenes 6 Superficie Construida 7 Superficie Útil 5 16. Superficie 29 Superficie Edificada 5 17. Comunidad 26 Comunidad de Propietarios 17 After obtaining these results, we proceeded with the second stage of our analysis, separating the five moves in each of the sample texts. In the English corpus, 8 samplesincludedtheCommencementmove,4thePreamble,10theClausemove,9the Testimonium clause move and 1 the Schedules. As for the Spanish part, 15 samples included the Commencement, 18 the Preamble and the Clause moves, 17 the Testimoniumclause move, and5samplesincludedSchedules. We carried out the same analysis in the whole corpus, that is to say we identified the keywords in each of the moves with more than 25 occurrences. We found that only the Clause move included more than 25 occurrences for the keywords. Tables 6 and 7 show the keywords in the Clause move in English and Spanish respectively. Table 6. Key words in the English Clause move Key words Occurrences Key words Occurrences 1. Seller(s) 209+030 15. Time 39 2. Buyer(s) 158+36 16. Money 35 3. Agreement 110 17. Days 34 4. Property 110 18. Estate 66 5. Closing* 93 19. Instalment 30 6. Purchaser 92 20. Parties 30 7. Title 88 21. Amount 28 8. Contract 66 22. Terms 27 9. Rights 64 23. Conditions 25 10. Date 58 24. Sale 26 11. Purchase 48 25. Payment 25 12. Price 41 26. Earnest 26* 13. Deed 40 27. Tax(es) 25 14. Mortgage 40 28. Deposit 25 Lexical chunks in English and Spanish sales contracts 47 © 2019. John Benjamins Publishing Company All rights reserved Table 7. Key words in the Spanish Clause move Key words Occurrences 1. Parte 465 2. Precio 135 3. Contrato 134 4. Compraventa 123 5. Gastos 114 6. Escritura 113 7. Pago 108 8. Préstamo 88 9. Entrega 83 10. Intereses 61 11. Propiedad 61 12. Tipo 50 13. Otorgamiento 41 14. Impuesto 37 15. Requerimiento 35 18 keywords, out of 28 formed lexical chunks in English, and 14 out of 15 in Spanish; seller, buyer, purchaser, purchase, days, installment, parties and sale did not form any lexical chunks in English, while in Spanish compraventa was the only keyword which did not form any lexical chunk. Keywords that did not appear in lexical chunks were discarded. See Tables 8 and 9 for the lexical chunks found in both corpora. Table 8. Lexical chunks in the English Clause move Key word Lexical chunks Occurrences Security agreement 3 Entire agreement 2 Acceptable agreement 1 Completed agreement 1 Implied agreement 1 Sales agreement 1 Verbal agreement 1 1. Agreement (s) 110+35 Written agreement 1 Personal property 10 Adjacent property 2 Adjoining property 2 2. Property 110 Real property 1 48 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved Table 8. (continued) Key word Lexical chunks Occurrences Real estate property 1 Residential property 1 Closing date/day of closing 6+53. Closing 93 Closing time/time of closing 5+4 Closing of title 6 Marketable title 5 Affidavit of title 2 4. Title 88 Abstract of title 1 Residential contract 6 Timber contract 3 Sole contract 2 Entire contract 2 5. Contract 64 Final contract 1 Closing date 5 Effective date 2 Possession date 1 6. Date 58 Time reference date 1 Homestead rights 2 Purchaser’s rights 1 7. Right(s) 64 Seller’s rights 1 Purchase price 32 Total purchase price 2 8. Price 41 Sale price 2 Warranty deed 12 Trust deed (purchase money mortgage) F 10 General mortgage deed 3 9. Deed 40 Sale deed 1 Existing mortgage 410. Mortgage 40 Purchase money mortgage (trust deed) 3 Earnest money 36 Purchase money 5 11. Money 35 Part purchase money 1 Real estate 4812. Estate 66 Unit ownership estate 1 13. Instalment 30 Monthly instalments 2 14. Amount 28 Full amount 2 15. Terms 27/ conditions 25 Terms and conditions 9 Lexical chunks in English and Spanish sales contracts 49 © 2019. John Benjamins Publishing Company All rights reserved Table 8. (continued) Key word Lexical chunks Occurrences Additional terms and conditions 3 Down payment 6 Final payment 3 Monthly payment 2 Installment payments 2 Monthly installment payments 2 16. Payment 25 Interest payment 1 Real estate taxes 8 Property taxes 5 General taxes 3 Special taxes 3 Transfer taxes 2 County transfer taxes 2 Ascertainable taxes 1 17. Taxes 25 Federal taxes 1 Money deposit 3 Buyer’s deposit 2 18. Deposit 25 Earnest money deposit 2 Table 9. Lexical chunks in the Spanish Clause move Key word Lexical chunks Occurrences Parte Compradora 250 Parte Vendedora 148 Parte de Notaría 6 Parte Expositiva 5 1. Parte 456 Parte Proporcional 5 Precio de la Compraventa 10 Precio de la Vivienda 9 2. Precio 24 Precio Pactado 5 3. Contrato 9 Contrato de Compraventa 9 Gastos e Impuestos 14 Gastos Comunes 11 Gastos de Gestión 6 4. Gastos 36 Gastos Comunes de Propiedad Horizontal 5 Escritura Pública 435. Escritura 93 Escritura de Compraventa 25 50 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved Table 9. (continued) Key word Lexical chunks Occurrences Escritura Pública de Compraventa 17 Escritura de Carta de Pago 5 Escritura de Préstamo 5 Pago del Precio Aplazado 11 Pago De Contribuciones, Arbitrios e Impuestos 6 6. Pago 23 Pago de Gastos Comunes De Propiedad Horizontal 6 7. Propiedad 22 Propiedad Horizontal 22 Préstamo Hipotecario 18 Préstamo Directo 7 Préstamo con Garantía Hipotecaria 5 8. Préstamo 35 Préstamo Cualificado 5 Entrega de Llaves 18 Entrega de la Vivienda 16 9. Entrega 40 Entrega de Llaves y Toma de Posesión 6 10. Intereses 7 Intereses de Demora 7 Tipo de Conversión 9 Tipo de Interés 5 Tipo Aplicable 3 11. Tipo 20 Tipo Impositivo 3 12. Otorgamiento 20 Otorgamiento de la Escritura 20 Impuesto Sobre el Valor Añadido 8 Impuesto Municipal sobre el Incremento del Valor de los Terrenos de Naturaleza Urbana 6 Impuesto(s) sobre Transmisiones y Actos Jurídicos Documentados 7 13. Impuesto(s) 24 Impuesto de Actos Jurídicos Documentados 3 Requerimiento Fehaciente 514. Requerimiento 8 Requerimiento Notarial o Judicial 3 Finally, Table 10 shows the raw results of our selection and classification of both lists of keywords and their concordance lines. Only those keywords occurring in both languages were retained for the study. Each of the keywords is dealt with separately, with the total number of lexical chunks identified for each indicated in parentheses after the keyword. 10 Keywords in English and 8 in Spanish were retained for the study, and a total of 19 lexical chunks in English and 13 in Spanish were analyzed. Lexical chunks in English and Spanish sales contracts 51 © 2019. John Benjamins Publishing Company All rights reserved Table 10. Keywords occurring in English and Spanish in our corpus of real property sales contracts Keyword Freq Lexical chunks Keyword Freq Lexical chunks Precio 135 Precio de la compraventa 10 Precio de la vivienda 9 Price/ money 41+35=76 (price/ money) Purchase price 32 Sale price 2 Total purchase price 2 Purchase money 5 Contrato 134 Contrato de compraventa 9 Agreement/ contract 145+64=209 Agreement/ contract Sales agreement 1 Final contract 1 Gastos 114 Gastos e impuestos 14 Costs 7 Closing costs 6 Escritura 113 Escritura pública 43 Escritura de compraventa 25 Escritura pública de compraventa 17 Deed 40 Trust deed (= purchase money mortgage)10 Sale deed 1 Préstamo 88 Préstamo hipotecario 18 Mortgage 40 Purchase money mortgage 3 Propiedad 61 Propiedad horizontal 22 Property 110 Real property 1 Real estate property 1 Residential property 1 Otorgamiento 41 Otorgamiento de la escritura 20 Closing 12 Closing title 6 Closing of title 6 Impuesto(s) 72 Impuesto municipal sobre el incremento del valor de los terrenos de naturaleza urbana 6 Impuesto(s) sobre transmisiones y actos jurídicos documentados 7 Impuesto de actos jurídicos documentados 3 Tax(es) 25 Real estate taxes 8 Property taxes 5 General taxes 3 Transfer taxes 2 52 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved 5. Analysis and discussion of results The lexical chunks identified above give rise to a number of observations, from the point of view of both form and meaning. We will mention aspects that are common to both English and Spanish and highlight differences when required. 1. Two-word units are far more common than larger units (cf. Closing costs -6 occurrencesvs. Closing will be held be on or about -2 occurrences-; escritura pública -43 occurrencesvs. escritura pública de compraventa -17 occurrences). In terms of form, this pattern corresponds to Gläser’s (1994/1995) word-like phrases, which she determined to be the most common in LSP, what is confirmed by our findings. 2. Generally, the two-word units consist of the keyword along with a descriptor. The descriptor is normally a noun used adjectivally in English (e.g. trust deed -10 occurrences-) or a prepositional phrase in both languages (e.g. closing of title -6 occurrences-; contrato de compraventa -9 occurrences). In English, we also found a preposition+noun unit (at closing), but this combination was not found in Spanish. These structures correspond to what Biel (2014) classified as term-forming patterns. Given both the nature and the style of our corpus (real property sales contracts), the trend of two-word units formed by Adj+N (inverse order in Spanish) or N+(Prep/Cnj)+N could be expected: thepurpose ofa salescontract isto agreeuponthe termsandconditions ofthe property to be sold, and the obvious way to do that is to combine descriptors (adjectives or nouns used adjectivally) with the property itself or one of its key aspects (the terms of the contract), represented by the key terms of the genre. The fact that the headwords of all these chunks are key terms of the genre and the absence of arbitrariness in the construction of these patterns demonstrates that the chunks found belong to the terminology of the field (Sager 1997). 3. A typical feature of our two-word units in English is that the collocate of a given keyword is also used as a collocate of another keyword. In other words, the same collocate appears in two or more lexical chunks. This is the basic formation principle for term-forming patterns according to Biel (2014). Although it is attested in the Spanish corpus, this pattern is not a prevalent feature in this language. The following examples show collocates that combine with one keyword and the keywords they collocate with: Closing →title, date, time, costs, agent / Compraventa →contrato, precio, escritura. Although the keywords are different collocates are adjectival nouns, the meaning of descriptors does not vary. In term formation, this strategy is called compounding, and “serve(s) the purpose of closer determination of a concept – narrowing Lexical chunks in English and Spanish sales contracts 53 © 2019. John Benjamins Publishing Company All rights reserved its intension – while at the same time showing the relation that exists between the new concept and its origin.” (Sager 1997,30) 4. Very few lexical chunks are restricted semantically in our corpus. This is also in line with Biel’s (2014) observations: the meaning of most of the lexical chunks is simply made up of the sum of the meanings of their parts. This results in “transparent terms [that attempt] to overcome the arbitrariness of natural language designation” (Sager 1997,26), leading to a better understanding on the users’ part. Our lexical chunks in both English and Spanish are semantically transparent, rather than opaque, for the most part, e.g. transfer taxes / préstamo hipotecario. 5. Some keywords form two and three word units that hold a semantic relationship: the prevalent relation in English is synonymy, while in Spanish, it is hyperonymy-hyponymy. An example of the former is the case of Real Estate Taxes and Property Taxes, which are considered synonyms in English according to the following definition: “Property tax is a tax assessed on real estate”.4 An example of the latter is the case of Escritura pública (deed) and Escritura pública de compraventa (Sale deed) in Spanish, the second being a type of deed, thus a hyponym; according to www.housel.com, a sale deed is “una escritura firmada ante notario para la posterior inscripción del inmueble en el Registro de la Propiedad” (a deed signed with a notary so as the real property can be registered at the Registrar Office; our translation). 6. Regardless of semantic relationships, smaller lexical chunks are often combined to form larger units as in the following examples: Real property and Real estate property;Purchase price and Total purchase price in English, and escritura pública and escritura pública de compraventa in Spanish. 7. In English, Adj / Noun used adjectivally+Noun units (e.g. purchase price) can also be expressed as Noun+Linking verb+Adj units (the price will be paid as follows). According to Biel’s classification (2014), we could identify these as transformations of term-forming patterns into term-embedding collocations. As we consider the latter a syntactic variation of the former, we have not listed it as a separate unit above. Spanish does not resort to this strategy according to the corpus. 8. Verbs used in real property sales contracts are generally of the linking variety (to be;ser) or are very general (to hold; efectuar). Frequently these verbs are used with auxiliary verbs implying obligation (in the active or passive voice). Thiscan be explainedbythe fact thatsales contracts imply binding obligations for the parties. For instance, sentences such as the following, forming lexical chunks, are very common in our corpus: 4. Investopedia. Property tax (https://www.investopedia.com/terms/p/propertytax.asp). Accessed 15 March 2019] 54 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved This agreement shall become effective Closing will be held on------- day of----- This Agreement shall be null and void […] renunciando a efectuar cualquier reclamación Los gastos e impuestos que se deriven de dicho otorgamiento serán satisfechos por las partes con arreglo a Ley--- These examples are what Biel described as “genre-specific recurrent grammatical patterns” (2014,179). In our case, they are formed mostly by performative verbs, “the main verb in a sentence whose literal utterance in appropriate circumstances constitutes the performance of an illocutionary act named by an expression in that very sentence in virtue of the occurrence of that expression.” (Searle 1989,537) This specific characteristic found in our corpus was also expectable, as performative verbs are a common feature in contracts, both in English and Spanish (Alcaraz and Hughes 2002; Alcaraz et al. 2002). 6. Conclusion Our description and classification of lexical chunks in a bilingual corpus of real property sales contracts in English and Spanish has shown that the language of real property sales contracts contains a fairly large number of such units: 19 lexical chunks involving 10 keywords in English, and 13 lexical chunks involving 8 keywords in Spanish. There seem to be more recurring combinations in English than in Spanish, although the occurrence of combinations is generally higher in English; this could perhaps be explained by the fact that there is a greater use of synonyms in English than in Spanish, which would result in fewer occurrences in Spanish of the same collocate with a given keyword. Our methodology allowed to show that the preferred form of the lexical chunks identified is word-like phrases (Gläser 1994/1995). The majority of these units, both in English and in Spanish, are term forming patterns, which agrees with Biel’s study (2014), consisting of the keyword and a descriptor in the form of an adjective or noun used adjectivally, preposition+noun or noun+preposition+noun. However, we also found some term-embedding collocations (Noun+linking verb+adjective) as the price will be paid as follows and grammatical patterns such as Closing will be held on------- day of----- or Los gastos e impuestos que se deriven de dicho otorgamiento serán satisfechos por las partes con arreglo a Ley--- Most lexical chunks forming term patterns in English and Spanish are fixed syntactically or semantically (Biel 2014,180). However, as we stated before (see Lexical chunks in English and Spanish sales contracts 55 © 2019. John Benjamins Publishing Company All rights reserved Sections 2and3andabove),ourcorpusshowsthattheypresentdifferentdegreesof stability.AmongtheEnglishlexicalchunksretainedforthisstudy,15canbeconsidered fixed such as Real estate property,Property taxes or Closing costs. At the other end,thatistosay,lexicalchunkswithaloosedegreeofstability,wefoundFinalcontract.Finally,3aretobedefinedwithamid-degreeofstability:Generaltaxes,Transfer taxes and Purchase money. In Spanish, 7 lexical chunks can be considered fixed such as Impuesto Municipal sobre el Incremento del Valor de los Terrenos de Naturaleza Urbana or propiedad horizontal. Additionally, there are 3 which could be consideredto havea mid-degreestability, as is thecase ofotorgamiento de la escritura.Lastly,3havealoosedegreeofstability,e.g.gastoseimpuestos. Most units are semantically transparent, with both the collocate and the keyword retaining an individual meaning and the meaning of the units being the sum of the meanings of their parts, which again agrees with Biel’s description of termforming patterns (2014,180). Our findings are based on a given genre in the field of contracts: real property sales contracts and using a small corpus. It would be interesting to see if they remain valid if a bigger corpora was used or if another genre within sales contracts, that of personal property for example, were analyzed. That is a study we intend to undertake in the future. Acknowledgements The authors belong to the ACTRES (Análisis Contrastivo y Traducción Especializada) research group. This paper has been written within the research project “Proyecto de investigación Producción textual bilingue semiautomatica inglés-espanol con lenguajes controlados: parametrización del conocimiento experto para su desarrollo en aplicaciones web 2.0 y 3.0”, supported financially by Ministerio de Ciencia e Innovación in 2016 (FFI2016-75672-R). References Alcaraz, Enrique and Brian Hughes. 2002. El Español Jurídico. Barcelona: Ariel. Alcaraz, Enrique, Miguel Ángel Campos, and Cynthia Miguélez. 2007. El Inglés Jurídico Norteamericano [North-American Legal English]. Barcelona: Ariel. Biber, Douglas. 1993. “Representativeness in Corpus Design.” Literary and Linguistic Computing 8 (4): 243–257. https://doi.org/10.1093/llc/8.4.243 Biber, Douglas, Susan Conrad, and Randy Reppen. 1999. Corpus Linguistics: Investigating Language Structure and Use. Amsterdam: John Benjamins. Biber, Douglas, Ulla Connor, and Thomas A. Upton. (eds). 2007. Discourse on the Move. Using Corpus Analysis to Describe Discourse Structure. Amsterdam: John Benjamins. https://doi.org/10.1075/scl.28 56 Belén López Arroyo & Leticia Moreno Pérez © 2019. John Benjamins Publishing Company All rights reserved