miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
43
Abstract
This paper aims to analyse the compositional argument selection process
represented by different syntactic alternations within the specialised domain of
cooking, thus contributing to the characterisation of this specialised discourse.
The syntactic alternations studied include canonical actives, passives, causative/
inchoative alternations, middles and Instrument-subject alternations. These
constructions allow the incorporation of cooking verbs (Levin 1993) and denote
divergent argument structure realisations. As indicated here through
compositional analysis, the constructions contain distinctive N+V qualia pairs. As
in Pustejovsky (1991, 1995), this paper follows a lexico-semantic approach and
applies a corpus-based methodology to examine and compare over 8,300
contextualised examples from two corpora (a specialised corpus on cooking and
a general corpus) using the Sketch Engine corpus tool. The results show that the
syntactic alternations examined follow related but distinctive underlying patterns
in semantic composition, and thus are construed with N+V qualia pairs that
characterise the specialised discourse of cooking.
Keywords: argument selection, compositional analysis, syntactic alternations,
discourse of cooking, qualia pairs.
MACARENA PALMA GUTIÉRREZ
Universidad de Córdoba
l82pagum@uco.es
<https://orcid.org/0000-0002-0558-9358>
COMPOSITIONAL ARGUMENT SELECTION IN N+V
QUALIA PAIRS WITHIN THE DISCOURSE
OF COOKING: A CORPUS-BASED STUDY
SELECCIÓN ARGUMENTAL COMPOSICIONAL
EN PARES DE QUALIA N+V EN EL DISCURSO
DE LA COCINA: UN ESTUDIO BASADO EN CORPUS
https://doi.org/10.26754/ojs_misc/mj.202510185
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
44
Resumen
El objetivo de este trabajo es analizar el proceso de selección argumental
composicional representado por diferentes alternancias sintácticas en el ámbito
especializado de la cocina y, de este modo, contribuir a la caracterización de este
discurso especializado. Las alternancias sintácticas objeto de estudio son las activas
canónicas, las pasivas, las alternancias causativo/incoativo, las medias, y las
alternancias de sujeto instrumento. Estas construcciones permiten la incorporación
de verbos de cocinar (Levin 1993) y denotan realizaciones divergentes de la
estructura argumental y, en consecuencia, como se atestigua aquí, contienen pares
de qualia N+V distintivos en el análisis composicional. En línea con Pustejovsky
(1991, 1995), este trabajo sigue un enfoque léxico-semántico y una metodología
basada en corpus para analizar y comparar 8300+ ejemplos contextualizados de
dos corpus (un corpus especializado de cocina y un corpus genérico) mediante el
uso del software Sketch Engine. Los resultados muestran que las alternancias
sintácticas examinadas siguen patrones subyacentes relacionados pero distintivos
en la composición semántica y, por lo tanto, se interpretan con pares de qualia N+V
que caracterizan el discurso especializado de cocina.
Palabras clave: selección argumental, análisis composicional, alternancias
sintácticas, discurso de la cocina, pares de qualia.
1. Introduction
The language of cooking has been widely explored from a linguistic perspective
(e.g. Lévi-Strauss 1966; Lehrer 1969, 1972; Newman 1975; Bator 2014). This
paper contributes to the characterisation of this specialised domain by contrasting
data from two corpora: a specialised corpus of cooking and a general corpus of
English. Particularly, it focuses on the usage of N+V configurations and the most
productive syntactic patterns typically found in this specialised domain,
contrasting these patterns with those from the general corpus.
Scholars like Casademont (2014) and Durán and L’Homme (2020: 37) consider
verbs as ‘conveyances of knowledge’ that help characterise specialised discourse
because they specify information about argument structure in their corresponding
cultural domains. Unlike purely verb- or noun-centred approaches to
compositionality (Sager 1990; Hale and Keyser 2002), this study follows
Pustejovsky’s (1991, 1995) ideas, thus advocating a lexico-semantic approach
that spreads the semantic load across all the constituents of the utterance. Nouns
and verbs are both considered specialised units of language in specific domains
and as such contribute to the syntactic and lexico-semantic characterisation of a
specialised discourse.
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
45
This study identifies typically-occurring N+V combinations in the discourse
of cooking as qualia pairs, that is, as linguistic elements that are paired
depending on the information predicated by a given verb about the meaning
of a particular noun. This linguistic connection, in fact, is triggered by our
basic knowledge about the nominal entity in question and our conceptualisation
of it in terms of its more inherent features, known as qualia roles (Pustejovsky
1991, 1995). A N+V qualia pair is “a combination in which the predicate
expresses one of the qualia values of the noun (like picture-paint, book-read, or
house-build)” (Pustejovsky and Jezek 2016: 13). The qualia relational structure
thus involves those lexico-semantic, syntactic and conceptual constraints that
are “based on the idea that there is a system of relations that characterises the
semantics of nominals” and “serves to specify the reading of a verb” (Yoshimura
1998: 115).
The hypothesis of this study is that there exists a correlation between the most
productive N+V qualia pairs and the most typically-occurring syntactic structures
in a given specialised domain, in contrast to those (qualia pairs and syntactic
structures) found in a general corpus. Therefore, there should be a correspondence
between the lexico-semantic and the syntactic features that characterise the
specialised discourse of cooking. For example, if corpus data confirms the N+V
qualia pair chef _cook as significantly productive, then it would follow that a
syntactic canonical transitive structure with an agentive subject and a patientive
object could be frequently found in this specialised discourse, thus revealing a
given pattern of qualia structure that specifies the meaning of the noun in relation
to the semantics of the verb.
This paper is organised as follows. Section 2 presents the analytical tools used
(namely, syntactic alternations with cooking verbs and notions of qualia and co-
specification phenomena). Section 3 describes the methodology employed.
Section 4 presents the main findings and a discussion of the results. Finally,
Section 5 offers some closing remarks.
2. Tools for Analysis
In this paper I examine the interaction between the lexico-semantic and syntactic
features that characterise the discourse of cooking. I analyse the frequency of
occurrence of different grammatical patterns and the most productive N+V qualia
pairs. To do so, Subsection 2.1 explores the different syntactic alternations that
appear with cooking verbs, and Subsection 2.2 examines the main principles of
qualia structure and co-specification phenomena.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
46
2.1. Syntactic Alternations with Cooking Verbs
Placing a particular nominal entity in subject position is anything but random,
since doing so requires a process of lexico-semantic and discourse-pragmatic
profiling. As stated in Palma Gutiérrez, “profiling is related to the specific
portrayal of the foregrounded domain of a given linguistic expression as the focus
of attention in discourse” (2024: 139). Depending on the profiling/defocusing
phenomena involved in each case, distinct portions of the action chain and
argument structure realisations are represented. This leads to the configuration
of different syntactic patterns where diverse energetic interactions occur among
the participants, that is, Agent, Patient and Instrument (Langacker 2013).
According to Levin’s (1993) typology, cooking verbs are classified by the distinct
methods or techniques of cooking they describe, such as baking, frying or
boiling. Their prominence and frequent use in the specialised discourse of
cooking reflects the centrality of these actions to this domain (cf. Levin 1993:
244). Following Levin’s classification, these verbs participate in the syntactic
patterns illustrated in Examples 1-5:1
(1) Jennifer baked the potatoes (with her new oven).
(2) The potatoes were baked (by the chef).
(3) The potatoes baked.
(4) Idaho potatoes bake beautifully.
(5) This oven bakes potatoes well. (Adapted from Levin 1993: 243-244)
The basic/canonical active form in Example 1 follows the SVO syntactic pattern:
it contains a +Animate agentive subject (Jennifer) which is profiled syntactically
and a -Animate patientive object (the potatoes), which is defocused. It also contains
an oblique Instrument. Therefore, the flow of energy within the canonical action
chain follows the sequence Agent-Patient(-Instrument). Traditionally, the
transitive active clause is considered the most basic/unmarked syntactic pattern.
As found in corpus studies by Givón, the assumption of the higher productivity
of transitive actives “is associated, among other things, with the predication that
the unmarked member of a binary distinction […] is more frequent in text”
(1993: 52). Additionally, as Stockwell claims, “a prototypical subject acts as both
topic and agent, and alternative clause-patterns represent a deviation away from
this norm” (2002: 35). Accordingly, syntactic patterns with a different argument
structure alignment, such as those illustrated in Examples 2-5, are considered
syntactic alternations to the canonical pattern, as detailed below.
The passive construction in Example 2 portrays the same situation as the canonical
pattern, but with the order of arguments reversed. It profiles a ±Animate Patient
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
47
(the potatoes), and, when specified, the Agent occupies a defocused oblique object
position, introduced by a by-phrase (the chef). The causative/inchoative alternation
in Example 3 occurs with verbs of change of state/position to describe an eventive
situation (Levin 1993: 30). It follows the SV syntactic pattern since it only profiles
a -Animate patientive subject (the potatoes) and no Agent is coded. The middle
alternation in Example 4 follows the SVA syntactic pattern: it only profiles a
-Animate patientive subject (Idaho potatoes). It describes a potential/facilitative
(not an eventive) situation emphasised by the adverbial beautifully (26). Both
causatives/inchoatives and middles are intransitive counterparts of the basic/
canonical active form. In turn, the Instrument-subject alternation in Example 5 is
a transitive structure that follows the SVO syntactic pattern. However, it profiles
the oblique participant this oven (with an instrumental, not agentive, role) and a
patientive argument in object position (potatoes) (80). Here, the Instrument can
be understood as a metonymic extension of the Agent, paraphrased by the sequence
Agent-Patient-Instrument in Jennifer baked the potatoes with her new oven.
Following Pustejovsky’s ideas, patientive participants are classified by their
belonging to a natural or an artifactual kind (2001: 8; 2006: 54). The former are
described as naturally-occurring entities (e.g. water), whereas the latter are described
as artifacts, that is, objects created for a particular purpose (e.g. sandwich). Contrary
to artifactual objects, naturally-occurring entities lack an agentive value and denote
nominals that have not been created out of any intentional behaviour.
Table 1 summarises the features of the syntactic alternations examined in this paper
regarding the syntactic and semantic arrangement of their argument structures:
Canonical
actives
Passive
alternations
Causative/
inchoative
alternations
Middle
alternations
Instrument-
subject
alternations
Grammatical
roles at the
syntactic level
Subject and
Object
Subject (and
Oblique
Object) Subject Subject Subject and
Object
Subject’s
semantic role
(Profiled entity)
Agent Patient Patient Patient Instrument
Object’s
semantic role
(Defocused
entity)
Patient Agent X X Patient
±Animate
subject +Animate ±Animate -Animate -Animate -Animate
Table 1. Syntactic and semantic features of the alternations under examination
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
48
2.2. Qualia Structure and the Process of Co-specification
The theory of qualia structure (Pustejovsky 1991, 1995) establishes a mechanism
to represent lexical meaning based on a system of four dimensions of meaning,
called qualia, whose main function is to “capture different properties of objects,
as they are represented in language” (Pustejovsky and Jezek 2016: 3).
A single quale indicates a particular aspect of a word’s meaning through the
relationship between the concept expressed by the word and another concept
evoked by it. Qualia roles rely on the conceptual relations that a word may activate.
The four basic qualia roles are as follows:
• Formal qualia (Qf) encode taxonomic information about a lexical item
(type-of relations), and they answer these questions: ‘What type of thing is
this?’ and ‘What is its nature?’
• Constitutive qualia (Qc) focus on partonomic information about the
constituent parts/material of an object (part-of/made-of relations), and
they address these questions: ‘What are its constituent parts?’ and ‘What is
it made of?’
• Telic qualia (Qt) capture information about the purpose/function of an
entity (used-for/functions-as relations), and they answer these questions:
‘What is its purpose?’ and ‘How does it function?’
• Agentive qualia (Qa) refer to information about the origin of an object
(created-by relations), and they address these questions: ‘How did it come
into being?’ and ‘What brought it about?’
Figure 1 illustrates the qualia structure of the lexical item house in terms of its
qualia roles:
Figure 1. Representation of the qualia structure of the lexical item house (adapted from
Pustejovsky and Jezek 2016: 8-9)
Even though this view of lexical meaning is basically decompositional,2 this
model also examines compositionality, that is, “how a word meaning may or may
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
49
not compose with other meanings, and how it changes in the different contexts”
(Pustejovsky and Jezek 2016: 5). Consider the contextual modulation of word
meaning in Examples 6a- 6d below, paying special attention to how different
qualia roles are activated in the contexts provided:
(6) a. They have a three-story house.
b. Never forget to lock your house when you leave.
c. My cousin lives in a comfortable house.
d. It took four years to finish the house.
In 6a the lexical item house refers to a type of physical object, a building, thus
relying on a Qf relation between house and three-story. In 6b, the term house
metonymically evokes its most salient part (the door lock) in a Qc relation with the
predicate lock. In 6c the most salient feature of the house that is conceptually
activated in this context is its inhabitability and comfortability, that is, its function
(Qt). Finally, the conceptual relation between house and finish in 6d profiles a Qa
relation based on its process of creation/construction. Therefore, the meaning of
a lexical item adapts to the semantics of the elements surrounding it in a particular
context, thus profiling or activating the most salient features (qualia roles) evoked
in each case.
From this notion of contextual modulation of word meaning, it follows that
different elements in a grammatical construction can be paired in discourse in
accordance with their qualia structure to specify their meaning. In this study, I
concentrate on the relationship between nouns and verbs forming qualia pairs in
the specialised discourse of cooking.3 As proposed in Pustejovsky and Jezek, an
N+V qualia pair is a combination in which the verb promotes one of the qualia
values of the noun, as in book-read or house-build (2016: 13). This phenomenon is
known as co-specification. The results of this paper also show that metonymic
embedding can occur when certain qualia values are subsumed within others in
compositional analysis.
3. Data and Method
In this paper I conduct a corpus-based study of 8,385 instances to examine and
contrast the lexico-semantic and syntactic properties of four cooking verbs and
their syntactic alternations in a specialised corpus of cooking and a general corpus
(henceforth, SC and GC, respectively). I compiled 892 examples from the SC and
7,493 from the GC. Particularly, I examine those syntactic alternations that
involve productive N+V combinations in the domain of cooking.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
50
The data collection process and subsequent analysis was divided into different
phases. First, I used the Web search function of the Sketch Engine corpus tool
(Kilgarriff et al. 2004) to compile the SC using texts from the internet (e.g. food
blogs, cooking recipes, restaurant reviews). The resulting corpus contains
760,630 words. The GC used to contrast the data with the SC was the English
Web 2021 (enTenTen21) corpus, the largest English-language corpus available on
the platform (over 52 billion words).
The second step was to extract the most salient N+V qualia pairs in both corpora.
To do so, I used the Word Sketch function of Sketch Engine, which provides a list
of the most frequent collocates for a given target word in specific grammatical
relations. In this case, the target words were the four selected verbs (i.e. cook,
bake, boil, fry). I chose these four verbs for two reasons. First, as proposed by
Levin, these verbs “describe the basic methods of cooking”, and thus “are the
ones that show the widest range of properties” among their class (1993: 244).
The second reason was that these verbs proved to be highly frequent in the SC, as
demonstrated by the results displayed in the Wordlist function of Sketch Engine.
The Wordlist tool automatically generates frequency lists for the words in a
corpus. When a filter was applied to the SC, 2,241 items were found. In terms of
their frequency of occurrence, cook occupies the fourth position (3,273
occurrences), bake the ninth (2,243 occurrences), fry the twentieth (1,165
occurrences) and boil the twenty-third (1,100 occurrences).
I then manually coded the most frequent N+V collocates. Word Sketch collocates
are classified in terms of their association score with the target word4 and sorted
into categories depending on their grammatical relations. I analysed the collocates
in terms of the following two syntactic relations: ‘Words that serve as Subject of
the verb’ and ‘Words that serve as Object of the verb’. I selected the ten most
frequent collocates in each syntactic category in both corpora. As Sketch Engine
does not have the capacity to automatically filter out lexical/morphological
mismatches, these were discarded manually, removing the non-valid instances
that contained adjectives lemmatised as verbs (e.g. baking in baking soda) and
nominalised forms of verbs (e.g. fries in French fries).
Later, I sorted the Word Sketch results for both corpora by searching for those
nominals that co-occurred with the four cooking verbs and performed any of the
following semantic roles:
• Agents: [+Animate] entities relying on the value human
• Patients: [-Animate] entities associated with the value food
• Instruments: [-Animate] entities related to the value tool
Once the N+V combinations were retrieved, the contextualised instances in both
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
51
corpora were then analysed and sorted according to their syntactic alternations.
In the case of the GC, I analysed the first 100 examples from each N+V
combination.
4. Results and Discussion
In this section, I discuss the main results of the corpus analysis. Firstly, I present the
most salient N+V combinations from both corpora, distinguishing the semantic
role of the nominal entities in each case. After this, I examine the modulation of
word meaning in some N+V qualia pairs and their co-specified values at the
lexico-semantic level. Finally, I explore the qualia patterns in compositional
argument selection phenomena that were most syntactically productive in both
corpora.
Table 2 shows the Word Sketch instances that were selected and thus identified as
the most frequent N+V qualia pairs in both corpora. To compare these results,
the pairs are ordered in terms of the raw frequency of occurrence of the nominal
entities with each verb (Nº) and their normalised frequency in number of hits per
million tokens (Freq).5 The semantic roles of the nominal entities (whether Agent
(A), Patient (P) or Instrument (I)) and the total number of occurrences (in both
subject and object positions) in both corpora are also provided:
VERB COOK
Specialised corpus (SC) General corpus (GC)
Subject position Object position Subject position Object position
Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq)
potato(P) 6/(6.68) chicken(P) 98/(128) chef(A) 4,132/(0.07) food(P) 78,933/(1.28)
rice(P) 6/(6.68) potato(P) 66/(89.05) chicken(P) 768/(0.01) meal(P) 65,691/(1.07)
egg(P) 6/(6.68) fish(P) 48/(58.99) pasta(P) 757/(0.01) meat(P) 26,132/(0.42)
chicken(P) 5/(5.57) egg(P) 45/(53.44) meat(P) 751/(0.01) dinner(P) 25,159/(0.41)
fish(P) 3/(3.45) rice(P) 40/(52.71) rice(P) 734/(0.01) rice(P) 24,245/(0.39)
-- -- pasta(P) 35/(51.77) potato(P) 654/(0.01) dish(P) 21,247/(0.34)
-- -- food(P) 25/(34.51) oven(I) 537/(0.01) chicken(P) 19,254/(0.31)
-- -- corn(P) 23/(28.94) bean(P) 457/(0.01) breakfast(P) 13,998/(0.23)
-- -- onion(P) 21/(25.62) onion(P) 415/(0.01) vegetable(P) 12,525/(0.2)
-- -- noodles(P) 18/(20.04) steak(P) 374/(0.01) pasta(P) 9,741/(0.16)
Total nº 26 Total nº 419 Total nº 9,579 Total nº 296,925
TOTAL Nº: 445 TOTAL Nº: 306,504
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
52
VERB BAKE
Specialised corpus (SC) General corpus (GC)
Subject position Object position Subject position Object position
Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq)
oven(I) 6/(6.68) pie(P) 11/(12.24) bread(P) 5,696/(0.09) bread(P) 41,596/(0.68)
-- -- pastry(P) 8/(8.9) oven(I) 2,788/(0.05) dish(P) 32,684/(0.53)
-- -- fish(P) 8/(8.9) cookie(P) 2,567/(0.04) cake(P) 31,169/(0.51)
-- -- tart(P) 7/(7.79) cake(P) 2,437/(0.04) potato(P) 26,902/(0.44)
-- -- cake(P) 7/(7.79) pie(P) 1,410/(0.02) cookie(P) 22,328/(0.36)
-- -- bun(P) 6/(6.68) cupcake(P) 467/(0.01) bean(P) 20,099/(0.33)
-- -- cookie(P) 6/(6.68) pizza(P) 430/(0.01) pan(I*) 18,387/(0.3)
-- -- loaf(P) 6/(6.68) baker(A) 418/(0.01) pie(P) 10,299/(0.17)
-- -- chicken(P) 4/(4.45) pastry(P) 411/(0.01) apple(P) 5,040/(0.08)
-- -- biscuit(P) 3/(3.45) muffin(P) 322/(0.01) chicken(P) 4,373/(0.07)
Total nº 6 Total nº 66 Total nº 16,946 Total nº 212,877
TOTAL Nº: 72 TOTAL Nº: 229,823
VERB BOIL
Specialised corpus (SC) General corpus (GC)
Subject position Object position Subject position Object position
Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq)
water(P) 16/(17.81) potato(P) 29/(31.3) pot(I) 2,692/(0.04) water(P) 101,495/(1.65)
mixture(P) 6/(6.68) corn(P) 21/(23.62) kettle(I) 1,628/(0.03) egg(P) 26,026/(0.42)
-- -- mixture(P) 20/(22.26) water(P) 633/(0.01) potato(P) 10,461/(0.17)
-- -- turkey(P) 16/(17.81) mixture(P) 612/(0.01) rice(P) 4,823/(0.08)
-- -- cookie(P) 12/(13.36) liquid(P) 519/(0.01) milk(P) 3,117/(0.05)
-- -- ham(P) 9/(10.7) soup(P) 95/(0) pot(I*) 2,977/(0.05)
-- -- kettle(I*) 8/(8.9) wort(P) 95/(0) kettle(I*) 2,797/(0.05)
-- -- chicken(P) 6/(6.68) pasta(P) 87/(0) peanut(P) 1,776/(0.03)
-- -- wing(P) 6/(6.68) cook(A) 82/(0) pasta(P) 1,613/(0.03)
-- -- pierogi(P) 4/(4.45) potato(P) 65/(0) noodle(P) 1,390/(0.02)
Total nº 22 Total nº 131 Total nº 924 Total nº 156,475
TOTAL Nº: 153 TOTAL Nº: 157,399
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
53
Table 2. Frequency of the most salient N+V qualia pairs in both corpora
Table 2 illustrates the N+V combinations and lexico-semantic features that
characterise each corpus. With the verb cook, the most frequent semantic roles in
subject position are the Patient potato (in the SC) and the Agent chef (in the GC).
Additionally, both corpora reveal the saliency of Patients in object position
(chicken and food, respectively). The two corpora differ in that the verb cook only
occurs with Patients (in both subject and object positions) in the SC, whereas in
the GC, other semantic roles are found in subject position (chef as Agent and oven
as Instrument).
In the case of bake, the most salient semantic roles in subject position are the
instrumental participant oven (in the SC) and the patientive entity bread (in the
GC). No other nominal entities in subject position were found in the SC with the
verb bake, whereas in the GC, three different semantic roles were found in subject
position: Patient (bread), Instrument (oven) and Agent (baker). In both corpora,
all the participants have a patientive nature in object position, with the exception
of pan in the GC, working as Instrument and conveying a metonymic value. To
illustrate this, consider Example 7 from the GC:
VERB FRY
Specialised corpus (SC) General corpus (GC)
Subject position Object position Subject position Object position
Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq) Entity(Role) Nº/(Freq)
-- -- onion(P) 113/(125.7) egg(P) 246/(0) egg(P) 4,872/(0.08)
-- -- shallot(P) 18/(20.04) bacon(P) 161/(0) chicken(P) 3,815/(0.06)
-- -- chicken(P) 17/(18.92) onion(P) 114/(0) onion(P) 2,869/(0.05)
-- -- egg(P) 16/(17.81) garlic(P) 76/(0) bacon(P) 1,986/(0.03)
-- -- mushroom(P) 12/(13.36) cook(A) 74/(0) potato(P) 1,782/(0.03)
-- -- gnocchi(P) 11/(12.24) turkey(P) 72/(0) rice(P) 1,150/(0.02)
-- -- potato(P) 11/(12.24) sausage(P) 35/(0) turkey(P) 897/(0.01)
-- -- garlic(P) 10/(11.13) chicken(P) 33/(0) tortilla(P) 620/(0.01)
-- -- bacon(P) 8/(8.9) burger(P) 19/(0) noodles(P) 620/(0.01)
-- -- pancetta(P) 6/(6.68) tofu(P) 14/(0) tofu(P) 457/(0.01)
Total nº 0 Total nº 222 Total nº 623 Total nº 19,068
TOTAL Nº: 153 TOTAL Nº: 19,691
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
54
(7) Bake each pan 10-12 min.
Here, the term pan stands for the ingredients inside it, meaning ‘bake the
ingredients with the pan’. The conceptual metonymy underlying this process is
container for content.
Regarding boil, the most salient co-occurring semantic roles in subject position
refer to the Patient water (in the SC) and the Instrument pot (in the GC). Once
again, the GC presents the three different semantic roles found in subject position:
Instrument (pot and kettle), Patient (water) and Agent (cook). In the SC, the most
frequently co-occurring entities with boil have a patientive nature, either in subject
or object position. The only exception is kettle in object position (see Example 8
taken from the SC), which follows the same process as pan in Example 7:
(8) Boil the kettle.
Here, the term kettle stands for the water inside it, meaning ‘boil the water with
the kettle’. The conceptual metonymy underlying this process is container for
content as well.
In the case of fry, no statistically relevant nominal entities in subject position were
found in the SC, whereas two different semantic roles (Patient and Agent)
occurred in the GC, as illustrated by egg and cook, respectively. All the entities
found in both subject and object position in both corpora with fry have a
patientive nature (with the exception of cook in subject position in the SC).
The lack of nominal entities and variety of semantic roles in subject position
observed throughout the SC is due to the pervasive use of instructional
imperatives, where no agentive subject is syntactically coded, though semantically
recoverable as you. Consider the following instance in Example 9 taken from the
SC in this regard:
(9) Fry the onions.
Recipe texts commonly contain imperative structures that guide users in the
cooking process. The SC contains a higher concentration of this type of patterns
as compared with the GC. As detailed below in this section, the most productive
structure in both corpora is the basic/canonical pattern. The main difference is
that the GC tends to portray the whole action chain, including salient Agent and
Patient entities (in declarative patterns), whereas the SC focuses on patientive
participants (within imperative forms).
Below I discuss the corpus results by examining the subject-verb qualia pairs in
N+V combinations where we find the different semantic roles of the nominal
entity (Agent, Patient and Instrument). Figure 2 shows the qualia structure
representation of four lexical items and highlights the co-specified qualia values
of these nominal entities in subject position in combination with cooking verbs.
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
55
The four lexical items and their semantic roles are chef (as an Agent), bread (as
Patient of the artifactual kind), kettle (as Instrument) and water (as Patient of the
natural kind).
Figure 2. Qualia structure representation of nouns with different semantic roles + cooking verbs
As illustrated in Figure 2, when these four nominal entities co-occur with cooking
verbs, they profile different qualia values depending on their semantic roles
(Agent, Patient and Instrument when in subject position) and the most
conceptually salient information they provide. The list below captures the main
principles of argument selection of these lexical items to form N+V qualia pairs in
the specialised discourse of cooking:
• Agentive participants (like chef) + cooking verb (like cook) = [Qt = ø = Qit]
• Artifactual patientive participants (like bread) + cooking verb (like bake) =
[Qa]
• Naturally-occurring patientive participants (like water) + cooking verb (like
boil) = [Qf/Qc]6
• Instrumental participants (like kettle) + cooking verb (like boil) = [Qt]
In the case of agentive and instrumental participants with cooking verbs, different
types of telic values are denoted, as represented by chef and kettle in Figure 2
above. The former refers to the ‘indirect telic’ (Qit) and the latter refers to the
‘direct telic’ (Qt) value. According to Pustejovsky and Jezek (2016: 30), the (Qt)
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
56
characterises the entity as something used to perform a particular activity (as in
kettle_boil), whereas the (Qit) characterises the entity as something that has the
function of carrying out the action denoted (as in chef_cook).
Therefore, as stated in Pustejovsky and Jezek (2016: 29), even though the (Qt) is
mostly associated with instrumental objects (as in kettle_boil), a telic value is also
present in nouns that denote concepts such as professions (as in chef _cook),
agentive nominals (like runner_run) and functional locations (such as school_
learn).
As illustrated in Figure 2, chef denotes a person (Qf) having the ability (Qa) to
cook (Qit). Correspondingly, in the case of kettle, this nominal denotes a tool
(Qf) used to boil water (Qt). Hence, in N+V qualia pairs where the nominal entity
is an Agent, the cooking verb co-specifies the (Qit) of its argument (either in
subject or object position). Similarly, in those N+V qualia pairs where the nominal
entity is an Instrument, the cooking verb co-specifies the (Qt) of its argument,
either in subject or (oblique) object position.
The other N+V qualia pairs retrieved from the corpus analysis that also denote
the (Qit) relation between the Agent and a cooking verb are baker_bake, cook_boil
and cook_fry. These combinations only appear in the GC, since no salient Agents
were identified in the SC. However, a (Qt) relation is found in the following N+V
qualia pairs, where the nominal performs the semantic role of Instrument: oven _
bake (in both corpora), and oven_cook, oven_fry, utensil_cook and pan_bake (only
in the GC).
Regarding artifactual and naturally-occurring patientive participants with
cooking verbs, we observe different patientive entities (natural and artifactual
kinds) participating in cooking events, thus denoting divergent N+V qualia pairs.
This is represented by the lexical items bread and water, respectively, in Figure 2
above. Whereas artifactual Patient-oriented entities profile a (Qa) value, naturally-
occurring Patients rely on a (Qf/Qc) relation, as detailed below.
The patientive entity bread belongs to the artifactual kind since it has been
created intentionally through a baking event. In the field of cooking, artifactual
Patient-oriented entities denote specific types of food (Qf) that are meant to be
eaten (Qt) following a process of creation (Qa). Therefore, in N+V qualia pairs
where the nominal entity is an artifactual Patient, the cooking verb co-specifies
the (Qa) of its argument (either in subject or object position). The other most
productive N+V qualia pairs retrieved from the corpora analysis that also denote
artifactual patientive entities co-specifying a (Qa) value are potato _cook (in the
SC) and cake_bake and egg_fry (in the GC).
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
57
Finally, the patientive participant water is a naturally-occurring entity. As
represented in Figure 2, the default value of (nil Qa) of this type of entities
captures “the primacy of a natural origin” (Pustejovsky and Jezek 2016: 34),
since they have not been created through any activity or intentional behaviour.
The only patientive entity of a natural kind found in the corpus is water, profiling
a (Qf/Qc) relation with the predicate boil. Thus, water denotes a type of liquid
that can boil due to its internal composition (Qf subsuming Qc); it is meant to be
drunk (Qt); and its origin possesses a naturally-occurring nature (nil Qa).
Therefore, in N+V qualia pairs where the nominal entity is a naturally-occurring
Patient, the cooking verb co-specifies the (Qf/Qc) values of its argument (either
in subject or object position).
So far, I have analysed modulation of word meaning in N+V qualia pairs and their
co-specified values at a lexico-semantic level, paying special attention to salient
lexical items in the discourse of cooking and focusing on the semantic roles of
these nominal entities with regard to a set of cooking verbs. Let us now further
explore the qualia patterns in compositional argument selection phenomena at
the syntactic level. Examples 10, 11 and 17 were retrieved from the GC, whereas
Examples 12 -16 were taken from the SC:
(10) The chef had never cooked vegan food before.
(11) The cook boiled the water for sterilizing.
(12) The fish fillets cooked through after 10min.
(13) The water boiled after 3-4min.
(14) Baked potatoes cook in about half the usual time in an air fryer.
(15) Water boils rapidly.
(16) This oven baked the salmon recipe perfectly.
(17) The oven cooks quickly and evenly.
Examples 10 and 11 represent canonical active transitives, whereas Examples 12-
15 represent intransitive alternations classified as follows: 12 and 13 are instances
of the causative/inchoative alternation, respectively incorporating an artifactual
and a naturally-occurring entity as Patients in subject position (fish fillets and
water). Examples 14 and 15 represent middles, which also incorporate an
artifactual and a naturally-occurring entity as Patients in subject position (baked
potatoes and water). Finally, Examples 16 and 17 represent, respectively, the
transitive and intransitive counterparts of the Instrument-subject alternation
(with oven).
Examples 10-17 are represented in Figure 3 to profile their qualia patterns in
compositional analysis at the syntactic level.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
58
Figure 3. Qualia patterns in compositional analysis at the syntactic level
As shown in Figure 3, canonical actives within the field of cooking (as in
Example 10) are transitive constructions that typically consist of patterns in
which the cooking verb expresses the (Qit) of the subject. Depending on the
nature of the Patient (whether artifactual or naturally-occurring), the cooking
verb expresses a different pattern in qualia structure: in combination with
artifactual Patients (like vegan food), the cooking verb expresses the (Qa) of the
object, thus profiling the creation process that the entity undergoes. Alternatively,
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
59
in combination with a naturally-occurring entity (like water), the cooking verb
expresses the (Qf/Qc) value of the object, since no (Qa) is found in natural kinds.
Thus, canonical active structures with artifactual Patients involve a [Qit + Qa]
qualia pattern in compositional analysis, whereas canonical actives with naturally-
occurring entities as Patients profile the [Qit + Qf/Qc] qualia pattern.
Figure 3 further illustrates both causatives/inchoatives and middles as intransitive
alternations. They have a common syntactic one-argument structure with a
patientive subject. However, depending on the nature of the nominal entity
(whether artifactual or naturally-occurring), a different qualia pattern is profiled
in compositional analysis. In those structures with an artifactual Patient-oriented
subject (as illustrated in Examples 12 and 14, respectively), the cooking verb co-
specifies the (Qa) of the subject by conceptually implying its creation processes.
However, in those structures with a natural-kind Patient-oriented subject (as in
Examples 13 and 15, respectively), the cooking verb co-specifies the (Qf/Qc)
values of the subject, since no (Qa) is found in natural kinds. Thus, causatives/
inchoatives and middles with an artifactual Patient-oriented subject underlie the
[Qa] qualia pattern in compositional analysis, but with natural-kind Patient-
oriented subjects they undergo the [Qf/Qc] pattern. The main difference between
these constructions is that causatives/inchoatives involve a specific time reference,
whereas middles incorporate adverbial/modal modifiers (time-oriented adjuncts
in 14 and 15) that influence their aspectual properties and reinforce their non-
eventive nature (Palma Gutiérrez 2022: 44).
The same qualia analysis can be applied to the passive structure. Even though
Levin (1993) does not contemplate passives as possible alternations with cooking
verbs, this Patient-subject alternation has been found very frequently in the
corpora examined, as detailed below.7 The main difference between the passive
and the other Patient-oriented structures (middles and causatives/inchoatives) is
that the passive Agent is either defocused (in a by- clause) or omitted syntactically,
whereas in the other structures, the Agent is totally demoted.
Finally, the Instrument-subject alternations illustrated in Figure 3 capture both the
transitive and the intransitive syntactic counterparts (respectively shown in Examples
16 and 17). Semantically, both alternations contain Instruments in subject position
(oven), and thus, their cooking verbs express the (Qt) of these participants. The
main difference between these alternations is that the transitive counterpart
encodes the patientive object at the syntactic level, whereas the intransitive
counterpart conceptually evokes it via metonymy. As detailed below, in order to
analyse the intransitive counterpart, I examine the notion of conceptual
modulation of noun meaning (Pustejovsky and Jezek 2016: 12) based on the
predicate’s argument selection process through a metonymic operation.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
60
First, in the case of the transitive counterpart of the Instrument-subject alternation,
I observed the following. In the selectional context of the verb bake, the noun
salmon is used to explicitly denote the entity that undergoes the baking process
(Qa), thus profiling the [Qt + Qa] qualia pattern in compositional analysis.
However, in the intransitive counterpart, I explored the notion of conceptual
modulation of noun meaning via metonymy. In the selectional context of the verb
cook, the noun oven is used to implicitly denote the entity that undergoes the
cooking process (Qa) by means of the metonymic complex instrument for action
for result (Serrano-Losada 2015: 43). This metonymic complex implies a dual
analysis: first, a domain expansion metonymy whereby the instrumental entity
oven stands for the action denoted cooking (i.e. instrument for action), and then,
a domain reduction metonymy whereby the action of cooking stands for its resulting
product food (i.e. action for result). Therefore, intransitive alternations of the
Instrument-subject construction profile the [Qt[Qa]] qualia pattern in compositional
analysis, where the (Qa) value is metonymically embedded within the (Qt) of the
N+V qualia pair.
Yet, as observed in the corpora, another metonymic relation is possible with
intransitive Instrument-subject alternations like The kettle boiled. In this case, in
the selectional context of boil, the noun kettle is used to metonymically evoke a
container for content relation (‘kettle’ for ‘water contained in the kettle’). In
contrast to the analysis carried out in Example 17, where a (Qa) value is embedded
within the (Qt) relation between oven and cook ([Qt[Qa]]), in the case of The kettle
boiled, the embedded qualia values are (Qf/Qc) because the nominal water is of a
natural kind ([Qt[Qf/Qc]]). Exploring which type of nominal entity is most
frequently topicalised when in combination with these verbs demonstrates the
tendency of these N+V combinations to participate in certain syntactic alternations
more productively.
As shown in Table 3, the four verbs under study were more productive in the
canonical pattern in both corpora, despite their differences in size. The main
differences were found in the second most productive N+V pairs retrieved in
each case.
Specialised corpus (SC) General corpus (GC)
cook bake boil fry cook bake boil fry
Canonical 265 64 129 221 953 902 1042 975
Caus/Incho 20 0 18 0 661 499 436 500
Middle 2 0 4 0 139 296 208 38
Instr. Subj 2 6 0 0 99 213 206 0
Passive 156 2 2 1 115 69 32 110
Table 3. Frequency of the syntactic alternations with each verb in both corpora
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
61
The data displayed in Table 3 confirms the research hypothesis: there exists a
correlation between the most productive N+V qualia pairs and the most frequent
syntactic constructions found in the corpora. First, the N+V[cook] combinations
suggest the following: the most salient pairs are cook_chicken (in the SC) and
cook_food (in the GC), thus entailing a higher productivity of the basic/
canonical action chain with Patients in object position and Agents in subject
position (which can be elicited in the imperative form) in both corpora.
Additionally, other salient N+V[cook] combinations are potato_cook (in the SC)
and chef_cook (in the GC). The former indicates a higher productivity of
grammatical patterns containing patientive subjects (such as passives, causative/
inchoatives or middles) in the SC, whereas the latter implies a higher productivity
of Agent-subject structures, whether in the canonical form (i.e. also including a
Patient in object position) or in the unspecified object alternation (i.e. with no
object specified).
Second, the N+V[bake] combinations show that the most salient pairs are bake_
pie (in the SC) and bake_bread (in the GC), therefore pointing to a higher
productivity of the basic/canonical action chain with Patients in object position
in both corpora. Also relevant is the case of oven_bake, which is highly frequent
in both corpora and implies a higher productivity of the Instrument-subject
alternation (transitive and intransitive variants) in both corpora.
Third, the N+V[boil] combinations indicate that the most salient pairs are boil_
potato (in the SC) and boil_water (in the GC), thus revealing a higher productivity
of the canonical structure with Patients in object position in both corpora.
Additionally, the other most salient N+V[boil] combinations are water_boil (in
the SC) and pot _boil (in the GC). The former would lead to a higher productivity
of Patient-oriented subject (such as middles, passives or causative/inchoative
alternations) in the SC, whereas the latter would imply a higher productivity of
Instrument-subject alternations in the GC.
Finally, the N+V[fry] combinations show that the most salient pairs are fry_onion
(in the SC) and fry_egg (in the GC), therefore indicating a higher productivity of
canonical patterns with patientive objects in both corpora. Additionally, in the
GC we also observe another highly salient pair, egg_fry, therefore implying a
higher productivity of Patient-subject structures (like middles, passives or
causatives/inchoatives).
Therefore, the data provided in Table 3 above, together with the previous
discussion of the connections between the normalised frequency of occurrence of
certain N+V qualia pairs and certain syntactic patterns, demonstrate that these
correspondences are different in each corpus, and this contrast contributes to the
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
62
lexico-semantic and syntactic characterisation of the discourse of cooking. In
both corpora, the most frequent grammatical pattern is the canonical structure
with the four verbs. However, the second most salient syntactic alternations differ
in both corpora. In the GC, the second most productive structure is the causative/
inchoative alternation with the four verbs due to the saliency of N+V combinations
containing Patient-oriented subjects. On the other hand, in the SC, each verb is
more productive in different structures, either relying on the saliency of N+V
pairs with Patient- or Instrument-oriented subjects: cook is more frequent in the
passive form, bake in the Instrument-subject alternation, boil in the causative/
inchoative alternation and fry in the passive form.
5. Conclusions
This paper analyses naturally-occurring language by applying the principles of
argument selection phenomena to examine the motivating factors behind the
processes of qualia-pairing and co-specification in the specialised discourse of
cooking. To do so, I examined the selectional contexts of four cooking verbs
(cook, bake, boil and fry) and the most salient nominal entities (in subject/object
positions) in combination with these verbs in two corpora (a specialised corpus
and a general corpus). This led to the creation of different argument structure
realisations as well as distinctive syntactic alternations in which the N+V
combinations were found. The results of this corpus-based study shed light on
the lexico-semantic and syntactic characterisation of the specialised discourse of
cooking.
As demonstrated here, the type of nominal entity most frequently topicalised
with the cooking verbs examined shows the tendency of each N+V combination
to participate more productively in certain syntactic alternations. In both corpora,
the four verbs are more productive in the canonical pattern. Concerning the
remaining less prototypical syntactic alternations, there are different tendencies
depending on the nature of the most frequent N+V pairs with each verb. In the
GC, the most frequent non-prototypical alternation is the causative/inchoative
pattern with the four verbs. In contrast, in the SC, other grammatical structures
become salient.
Regarding the notion of qualia structure, I have also demonstrated that these
cooking verbs express a (Qit) value when combined with agentive subjects. This
is illustrated in the N+V qualia pairs chef_cook, baker _bake and cook_fry. These
cooking verbs express divergent patterns in qualia structure depending on the
nature of the patientive entities they accompany (whether artifactual or naturally-
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
63
occurring). Therefore, artifactual Patients (in both subject and object position)
with cooking verbs evoke a (Qa) relation in their qualia patterns. These nominals
co-specify the meaning of the predicates by conceptually implying their creation
process (Qa) and encoding the resulting outcome. This is illustrated in the N+V
qualia pair bread_bake, egg_fry and potato_cook. However, this relation is not
found in naturally-occurring Patients with cooking verbs, since natural kinds
have (nil Qa) and thus profile a (Qf/Qc) value in co-specification with the
predicates, as shown in the N+V qualia pair water_boil. Finally, the cooking verbs
examined here express a (Qt) value when combined with Instrument subjects in
both transitive and intransitive counterparts. This is illustrated in the N+V qualia
pairs pot_boil, oven_cook, oven_bake and kettle_boil.
When the above N+V qualia pairs are examined regarding the syntactic
alternations that these cooking verbs can undergo, the following patterns in
compositional co-specification are found:
(i) Canonical active structures represent one of these complex patterns
depending on the nature of the grammatical object: [Qit + Qa] or [Qit +
Qf/Qc].
(i) Causatives/Inchoatives, middles and passives illustrate one of these
simple patterns depending on the nature of the grammatical subject: [Qa]
or [Qf/Qc].
(iii) Instrument-subject alternations denote one of these qualia patterns
depending on the occurrence or not of a metonymically-based operation
on the grammatical subject: [Qt + Qa], [Qt[Qa]] or [Qt[Qf/Qc]].
This study has examined contextual modulation of word meaning by exploring
the process of compositional argument selection in N+V qualia pairs in the
specialised discourse of cooking. The significance of the present study for the
specialised domain of cooking is directly associated with its linguistic
characterisation in terms of its lexico-semantic and syntactic features, particularly,
when in combination with a usage-based approach. Future lines of research may
explore other syntactic alternations, as well as other metonymically-based
operations.
Acknowledgements
I would like to thank the anonymous reviewers for taking the time and effort to
review the manuscript. I sincerely appreciate all their valuable comments and
suggestions, which helped me in improving the quality of the paper.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
64
Notes
1. Levin also presents other
alternations with cooking verbs (adjectival
passive participle alternations (e.g. a baked
potato), and resultative phrases (e.g. Jennifer
baked the potatoes to a crisp)) (1993: 244).
However, these were discarded because the
former have a lexical (not a syntactic) nature,
and thus, no transitivity alternations can be
examined, and the latter, because no instances
were found in the corpora consulted.
2. The process of lexical
decomposition follows the idea that words
can be decomposed into semantic primitives
annotated as (±) binary values. For example, the
word chair can be decomposed as [-Animate],
[+Countable], [+Concrete], [+Artifact].
3. As Pustejovsky and Jezek
explain, “[a] qualia pair may take the form of a
verb-noun pairing, and adjective-noun
pairing, or a compound” (2016: 31). This paper
focuses exclusively on N+V qualia pairs to
analyse the argument selection process
within the specialised discourse of cooking.
4. Association score uses
pointwise mutual information between the
target word and its collocate, multiplied by
the log of the pair frequency for the particular
grammatical relation examined. Association
score uses the so-called logDice statistical
measure to automatically identify collocations
and frequent combinations of words and is
not affected by the size of the corpus (Rychlý
2008).
5. In Table 2, both raw frequency
and normalised frequency are provided. Even
though the raw frequency in the SC is minimal
compared to that of the GC, the normalised
frequency is higher in the SC. This is so
because the N+V pairs are more productive in
the SC than in the GC, despite their size.
6. The qualia representation of the
naturally-occurring entity water merges Qf
and Qc values because we cannot separate
what this entity is (Qf) from what it is made of
(Qc). In fact, we could specify the Qc value
since each water molecule is identical and is
made up of one oxygen atom and two
hydrogen atoms, chemically represented as
H2O. However, its Qf value cannot be
separated from this established Qc condition.
7. Two syntactic patterns not
proposed by Levin (1993) as potential
alternations with cooking verbs were
identified in the GC (not in the SC), particularly,
with the verbs cook and bake, distributed as
follows: 26 instances with cook and 19 with
bake in the unspecified object alternation, and
seven instances with cook and two with bake
in the benefactive alternation.
N+V Qualia Pairs within the Discourse of Cooking
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
65
Works Cited
BATOR, Magdalena. 2014. Culinary Verbs in Middle English. Peter Lang. <https://doi.
org/10.3726/978-3-653-04576-5>.
CASADEMONT, Anna Joan. 2014. “On the Elements Activating the Transmission of Specialised
Knowledge in Verbs”. Terminology 20 (1): 92-116. <https://doi.org/10.1075/term.20.1.05joa>.
DURÁN-MUñOZ, Isabel and Marie-Claude L’HOMME. 2020. “Diving into Adventure Tourism from a
Lexico-semantic Approach: An Analysis of English Motion Verbs”. Terminology 26 (1): 33-59.
<https://doi.org/10.1075/term.00041.dur>.
GIVóN, Talmy. 1993. English Grammar. A Function-based Introduction (Vol. II). John Benjamins.
<https://doi.org/10.1075/z.engram2>.
HALE, Ken and Samuel Jay KEYSER. 2002. Prolegomenon to a Theory of Argument Structure. The
MIT Press. <https://doi.org/10.7551/mitpress/5634.001.0001>.
KILGARRIFF, Adam, Pavel RYCHLý, Pavel SMRZ and David TUGwELL. 2004. “The Sketch Engine”. In
Williams, Geoffrey and Sandra Vessier (eds.) Proceedings of the Eleventh EURALEX International
Congress. Lorient, France: 105-116.
LANGACKER, Ronald. 2013. “Settings, Participants, and Grammatical Relations”. In Tsohatzidis,
Savas (ed.) Meanings and Prototypes: Studies in Linguistic Categorization. Routledge: 213-238.
LEHRER, Adrienne. 1969. “Semantic Cuisine”. Journal of Linguistics 5: 39-55. <https://doi.
org/10.1017/S0022226700002048>.
LEHRER, Adrienne. 1972. “Cooking Vocabularies and the Culinary Triangle of Lévi-Strauss”.
Anthropological Linguistics 14 (5): 155-171.
LÉVI-STRAUSS, Claude. 1966. The Culinary Triangle. Partisan Review.
LEVIN, Beth. 1993. English Verb Classes and Alternations: A Preliminary Investigation. The
University of Chicago Press.
NEwMAN, Aryeh. 1975. “A Semantic Analysis of English and Hebrew Cooking Terms”. Lingua 37:
53-79. <https://doi.org/10.1016/0024-3841(75)90004-2>.
PALMA GUTIÉRREZ, Macarena. 2022. A Family-resemblance Analysis of the Middle Construction: A
Functional-cognitive Approach. Phd Dissertation. Universidad de Córdoba.
PALMA GUTIÉRREZ, Macarena. 2024. “Profiling and Defocusing Phenomena in the Discourse of Fe/
male Novelists: A Corpus Based Approach”. In Jiménez-Navarro, Eva Lucía and Leonor Martínez
Serrano (eds.) Where Gender and Corpora Meet: New Insights into Discourse Analysis. Peter
Lang: 135-156.
PUSTEJOVSKY, James. 1991. “The Generative Lexicon”. Computational Linguistics 17 (4): 409-441.
PUSTEJOVSKY, James. 1995. The Generative Lexicon. The MIT Press. <https://doi.org/10.7551/
mitpress/3225.001.0001>.
PUSTEJOVSKY, James. 2001. “Type Construction and the Logic of Concepts”. In Bouillon, Pierrette
and Federica Busa (eds.) The Syntax of Word Meaning. Cambridge U.P.: 1-37.
PUSTEJOVSKY, James. 2006. “Type Theory and Lexical Decomposition”. Journal of Cognitive
Science 6: 39-76.
Macarena Palma Gutiérrez
miscelánea 71 (2025): pp. 43-66 ISSN: 1137-6368 e-ISSN: 2386-4834
66
PUSTEJOVSKY, James and Elisabetta JEZEK. 2016. A Guide to Generative Lexicon Theory. Oxford
U . P.
RYCHLý, Pavel. 2008. “A Lexicographer-friendly Association Score”. In Sojka, Petr and Aleš Horák
(eds.) Proceedings of Recent Advances in Slavonic Natural Language Processing. Masaryk
University: 6-9.
SAGER, Juan. 1990. A Practical Course in Terminology Processing. John Benjamins. <https://doi.
org/10.1075/z.44>.
SERRANO-LOSADA, Mario. 2015. “Multimodal Metaphorical and Metonymic Renderings of Pain in
Advertising: A Case Study”. Revista Electrónica de Lingüística Aplicada (RAEL) 1 (14): 35-50.
SKETCH ENGINE. 2023. Lexical Computing Limited. <https://www.sketchengine.eu/>. Accessed July
09, 2024.
STOCKwELL, Peter. 2002. Cognitive Poetics: An Introduction. Routledge.
YOSHIMURA, Kimihiro. 1998. “The Encyclopedic Structure of Nominals and Middle Expressions in
English”. Kobe Papers in Linguistics 1: 112-140.
Received: 02/02/2024
Accepted: 02/09/2024
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
