by Lucía Madrigal Acero
Visiting PhD fellow
We use collocations everyday without really noticing them. We wake up, take the bus to work and sit down at our desks. Sometimes, we are required to make decisions and take responsibility and, when we are exhausted, we need to take a break. After having lunch, we will work a few more hours and go home to our children doing homework. In the late evening, once we have taken notice that everything is alright, we will go to bed and have a good night’s sleep.
Collocations are a multi-faceted aspect of language. To begin with, they lie in the syntax-semantics interface, which accounts for their specific behaviour (Baños et al., 2022; Fendel, 2024). From a diachronic perspective, they undergo a continuous renewal which makes it complicated to trace their development over a long period of time (Baños, 2018). From a synchronic perspective, different kinds of collocations behave differently and obey different rules. So, what exactly is a collocation?
Collocations
Collocations are semi-lexicalised, semi-compositional and lexically restricted word combinations (Koike, 2001; Mel’čuk, 2023). Sometimes it is easier to define them by what they are not: they are neither compositional constructions ‒i.e. a regular phrase or sentence whose meaning is the result of adding the meaning of all its components‒ nor fixed phraseological units ‒i.e. non-compositional expressions that do not admit variation and whose meaning cannot be guessed by knowing the individual meaning of its words. Sometimes, this situation is represented as a continuum, where collocations lie in an intermediate position between compositional constructions and fixed phraseological units:
Table 1. Continuum from compositional constructions to fixed phraseological units
| Compositional constructions | Collocations | Fixed phraseological units |
| I had a chocolate cake. | I had a slice of cake. | The test was a piece of cake. |
| The nerves are responsible for carrying commands from the brain to other parts of the body. | He has the nerve to perform dangerous surgeries. | My little brother always gets to my nerves! |
| The heart is a vital organ. | He is the one who broke my heart. | I’m wearing my heart on my sleeve. |
Collocations can be classified according to the morphological category of the words involved (Koike, 2001). In the table above, there are collocations of noun + of + noun and verb + noun. In this post, I will be writing about verb + noun collocations.
Support-verb constructions
Support-verb constructions (SVCs) are a specific type of verb-noun collocations (Alonso Ramos, 2004; Jiménez López, 2016). Their main characteristics are (i) that the noun performs the function of the semantic nucleus of the phrase and is predicative, meaning that it has its own semantic arguments; and (ii) that verb functions as an operator, that is, it has the syntactic function of providing the necessary slots for the actualisation of the arguments of the noun in a sentence. The support verb is generally considered to be semantically empty, which is why the SVC can be substituted by a simplex verb without changing the meaning of the sentence (cf. ex. 1 and 2). In a way, the noun in the collocation works as the lexeme in the simplex verb, and the verb of the collocation has a function analogous to that of the verb endings in the simplex verb.
(1) καὶ νῦν, εἴ τις καὶ τότε ἐν τῷ πάσχειν οὐκ εἰκότως ὠργίζετό μοι, μετὰ τοῦ ἀληθοῦς σκοπῶν ἀναπειθέσθω (Th. 6.89.3)
‘And now, if anyone got unfairly angry with me back then because of my suffering, let this person be persuaded after examining it together with the truth’
(2) οἱ δὲ Λακεδαιμόνιοι ἀκούσαντες ὀργὴν μὲν φανερὰν οὐκ ἐποιοῦντο τοῖς Ἀθηναίοις (Th. 1.92.1)
‘The Lacedaemonians, after hearing this, did not get visibly angry at the Athenians’
Even though several linguistic theories have tried to disentangle the intricacies of collocations, the Meaning-Text theory is the one that targets phraseology and collocations more specifically (Mel’čuk et al., 1984; Polguère, 1998; Alonso Ramos, 2004; Mel’čuk, 2023). In the Meaning-Text theory, SVCs are described by means of simple lexical functions. A lexical function is the formal expression of the syntactic and semantic properties of the verb in a collocation. The lexical functions for support verbs are Oper ‒when the predicative noun appears as the direct object of the support verb (ex. 2)‒, Func ‒when the noun is the subject (ex. 3)‒, or Labor ‒where the noun appears as an oblique complement in a prepositional phrase (ex. 4) (retrieved from: https://dicogra.iatext.ulpgc.es/ [28/11/2024]).
(3) εἰ δέ τινες αὖ φοβοῦνται μὴ ματαία ἂν γένοιτο αὕτη ἡ κατασκευή, εἰ πόλεμος ἐγερθείη […] (X. Vect. 4.41.2)
‘If anyone fears that this plan should prove in vain, should war be stirred anew […]’
(4) ἡ μὲν γὰρ τὸν τῆς Ἑλένης γεγενημένον φορέσασα εἰς ἑταιρικὴν αἰσχύνην ἐνέπεσε (D.S. 16.64.2)
‘For one of them wore [the necklace] that once belonged to Helen and fell from grace into prostitution [lit. fell into the shame of prostitution]’
Other kinds of verb-noun collocations: a lexical conjugation
There are other kinds of verb-noun collocations where the verb is not entirely devoid of meaning. In the continuum between collocations and compositional constructions, these verb-noun collocations fall in an intermediate position, because the fact that the verb is not semantically empty makes them closer to compositional constructions.
One of these kinds of collocations are those with realizative verbs. A realizative verb is very similar to a prototypical support verb, but it is not entirely empty from a semantic perspective, since it expresses that the reason for the existence of the noun in a certain state of affairs is carried out. The typical example which is given in the literature is that of promise (Alonso Ramos, 2004), which is combined with two possible realizative verbs: to fulfill and to keep. It is also common to classify as such verbs that require an instrument, such as to cut with a saw or to drive a car (Mel’čuk, 2023). These instrumental nouns are not predicative, but can be reinterpreted as such in combination with certain verbs if we consider that an instrument is meant to be used by someone (an agent) for something else, in the case of saw to cut an object (a patient) and in the case of car to transport something (also a patient). Since these verbs are very similar to support verbs despite this semantic nuance, they are also described by means of simple lexical functions, which are syntactically equivalent to the ones used for support verbs: Real (≈Oper), Fact (≈Func) and Labreal (≈Labor).
There are also more complex collocations which involve an aspectual or diathetic change compared to the corresponding prototypical SVC. For instance, there is a clear contrast among the following examples:
(5) εἶπεν ὅτι οὐ πόλεμον ποιησόμενοι ἥκοιεν (X. An. 5.5.24)
‘He said that they were not coming to fight’
Oper1(πόλεμος) = ποιέομαι
(6) οὕτως καὶ τὸν πόλεμον τοῦτον διαλύσομεν (Ar. Lys. 569)
‘And so will we put an end to this war’
Liqu1Oper1(πόλεμος) = διαλύω
(7) πῶς χρὴ πρὸς τούτους ῥᾳδίως πόλεμον ἄρασθαι καὶ τίνι πιστεύσαντας ἀπαρασκεύους ἐπειχθῆναι; (Th. 1.80.3)
‘How can a war against them be undertaken so lightly and with what belief can we hurry an attack without being ready?’
IncepOper1(πόλεμος) = αἴρομαι
The example in (5) is an SVC and can be lexicographically described as shown in the gloss. Examples (6) and (7) are different. The verb διαλύω adds to the collocation the meaning of ‘putting an end to something’, which is glossed in the Meaning-Text theory with Liqu. The subindex 1 in Liqu1 indicates that the Cause is the first semantic argument of πόλεμος. When combined with Oper1, it requires a verb that has this meaning, takes the noun as a direct object and takes the first argument of πόλεμος as its subject-Cause. In (7), the verb αἴρομαι adds an inchoative value to πόλεμος, which is represented by Incep in the corresponding lexical function.
In the French academic tradition, these verbs which appear in verb-noun collocations, but are not support verbs, are called extensions à verbe support, or support-verb extensions in English, and they form a constellation of collocations that has sometimes been described as a lexical conjugation (Daladier, 1996; Gross, 1999; Tronci, 2009). The Meaning-Text theory describes them by means of complex lexical functions (IncepOper, CausOper, etc.), except in the case of realizative collocations, which, like SVCs, are described by means of simple lexical functions. The use of complex lexical functions implies a particular understanding of these collocations with support-verb extensions: the verb has the same syntactic functions as a support or a realizative verb, but it has an added semantic value that adds a layer of complexity to its interpretation.
The middle voice
Support verbs coincide in form with regular verbs, that is, for a support verb ποιέομαι there is also a semantically full ποιέω; for support verb τίθεμαι, there is also a full τίθημι, etc. The voice distribution is Classical Greek indicates when the verb is used as a semantically full verb or a support verb. For instance, ποιέω in the active is typically used in its prototypical meaning of ‘creation’ (8) or as a causative verb (9), whereas the middle voice indicates its behaviour as a support verb (10) (cf. Cock (1981) for the discussion of neutralisations in voice alternation):
(8) […] ἧχι ἑκάστῳ δῶμα περικλυτὸς ἀμφιγυήεις / Ἥφαιστος ποίησεν ἰδυίῃσι πραπίδεσσι (Il. 608)
‘[…] where the renowned crooked-legged Hephaistos had built a dwelling for each of them with the knowledge of his mind’
(9) ἢ καὶ τότε τοὺς ἀμύνεσθαι κελεύοντας πόλεμον ποιεῖν φήσομεν; (D. 10.61)
‘Or will we also say then that the ones who are asking to defend ourselves are causing a war?’
(10) = (5) εἶπεν ὅτι οὐ πόλεμον ποιησόμενοι ἥκοιεν (X. An. 5.5.24)
‘He said that they were not coming to fight’
However, in Postclassical Greek the middle voice is a receding category, meaning that it got progressively less productive, and, by the time of the redaction of the Septuagint, it is already frequent to find SVCs with ποιέω in the active (11).
(11) αἷμα εἰς πλῆθος ἐξέχεας καὶ πολέμους μεγάλους ἐποίησας (1 Ch. 22.8)
‘You have drained much blood from the masses and fought great wars’
This makes the interpretation of certain texts particularly tricky: before, the active voice was indicative of a full verb or a causative verb, whereas the middle gave away that the predicate had to be interpreted as an SVC. From this point onwards, ambiguity in certain cases becomes unavoidable.
Sociolinguistic variation
Linguistic variation and change in Postclassical Greek are inevitably marked by diglossia (Toufexis, 2008; Horrocks, 2014). In a case study, Anlauf (1960) proved that the Greek optative declined to the point of disappearing in hellenistic Greek only to be revived a few decades later due to Atticism. Atticism was a rhetorical and linguistic movement that reacted against language change in Postclassical Greek and performed as a conservative impulse for the preservation of Classical Greek or, in other words, for the avoidance of language change, which was considered as the decline of an otherwise perfect language.
Something similar to this death and revival of the optative, albeit probably more complex and definitely not equal, could be said to have happened to the middle voice. The Septuagint, and later the New Testament, were representative of the changes that the Greek language had undergone up until the moment when they were written. They reflect the gradual disappearance of the middle/active alternation, the middle voice being preserved mostly for media tantum verbs or very specific contexts. Middle ποιέομαι as a support verb is only rarely found in these texts. Around that time, Atticism revived Classical literature and started frowning upon these newer and, in their opinion, lesser works (Monaco, 2021). This imitation of literary models could be one of the reasons that, in later centuries and among educated authors such as Symeon Metaphrastes, who rewrote the Live of Daniel the Stylite in a higher level of style, middle ποιέομαι was still used in SVCs (12), alongside with active ποιέω in less elevated pieces of literature, such as the old Life of Daniel the Stylite (13).
(12) ἐντειλάμενος δὲ τοῖς αὐτοῦ πατράσιν ὁ τῆς μονῆς προεστὼς μὴ πυκνὰς ποιεῖσθαι πρὸς τὸν παῖδα τὰς προσελεύσεις, χαίροντας ἐκπέμπει γονεῖς τὸ καινότατον υἱοῦ στερομένους (DelStyl5 [10th c. AD])
‘The abbot of the monastery instructed the child’s parents not to visit him frequently and sent them off with joy, even though they were deprived of their extraordinary son’
(13) ἀναστὰς δὲ ὁ δοῦλος τοῦ Θεοῦ καὶ ποιήσας εὐχὴν λέγει πρὸς αὐτούς· «Ἰησοῦς ὁ Χριστὸς ὁ σωτήρ μου, εἰς ὃν ἐπίστευσα καὶ πιστεύω, […]» (DelStyl18 [5th c. AD])
‘The servant of God stood up, said a prayer and spoke to them: “Jesus Christ, my Saviour, in whom I have believed and believe […]”’
In fact, in a recent paper, I found a correlation between higher and later texts and the use of εὐχὴν ποιέομαι, and lower and older texts and the use of εὐχὴν ποιέω (Vives Cuesta & Madrigal Acero, 2022). The general situation with other collocations is more complicated because of the interaction between morphological voice, the different kinds of collocations that a verb may appear in, and the event structure of each noun and verb in a collocation. However, the sociolinguistic approach may soon cast some new light on the use of the middle voice in Postclassical Greek.
Bibliographical references
Alonso Ramos, M. (2004). Las construcciones con verbo de apoyo. Visor.
Anlauf, G. (1960). Standard late Greek oder Attizismus? : Eine Studie zum Optativgebrauch im nachklassischen Griechisch. Universität zu Köln.
Baños, J. M. (2018). Las construcciones con verbo soporte en latín: Una perspectiva diacrónica. Cahiers du Laboratoire de Recherche sur le Langage, 7, 21–51.
Baños, J. M., Tur, C., Jiménez López, M. D., & Jiménez Martínez, M. I. (Eds.). (2022). Collocations in theoretical and applied linguistics: From Classical languages to Romance languages. Guillermo Escolar/SEEC.
Cock, A. J. C. M. (1981). ΠΟΙΕΙΣΘΑΙ: ΠΟΙΕΙΝ. Sur les critères déterminant le choix entre l’actif ΠΟΙΕΙΝ et le moyen ΠΟΙΕΙΣΘΑΙ. Mnemosyne, 34(1), 1–62.
Daladier, A. (1996). Le rôle des verbes supports dans un système de conjugaison nominale et l’existence d’une voiz nominale en français. Langages, 121, 35–53.
Fendel, V. (Ed.). (2024). Support-verb constructions in the corpora of Greek: Between lexicon and grammar? Language Science Press. https://langsci-press.org/catalog/book/461
Gross, G. (1999). Verbes supports et conjugaison nominale. Revue d’études Francophones, 9, 70–92.
Horrocks, G. (2014). High Register Byzantine Greek: Diglossia and what lay behind it. In C. Carpinato (Ed.), Storia e storie della lingua greca (pp. 49–72). Ca’ Foscari.
Jiménez López, M. D. (2016). On Support Verb Constructions in Ancient Greek. Archivio Glottologico Italiano, 101(2), 180–204.
Koike, K. (2001). Colocaciones léxicas en el español actual: Estudio formal y léxico-semántico. Universidad de Alcalá/Takushoku University.
Mel’čuk, I. (2023). General phraseology: Theory and practice. John Benjamins. https://doi.org/10.1075/lis.36
Mel’čuk, I., Arbatchewsky-Jumarie, N., Iordanskaja, L., Mantha, S., & Polguère, A. (1984). Dictionnaire explicatif et combinatoire du français contemporain (1–4). Les presses universitaires de l’université de Montréal.
Monaco, C. (2021). The Origins and Development of Linguistic Atticism [Cambridge University]. https://doi.org/10.17863/CAM.72412
Polguère, A. (1998). La théorie Sens-Texte. Dialangue, 8–9, 9–30.
Toufexis, N. (2008). Diglossia and register variation in Medieval Greek. Byzantine and Modern Greek Studies, 32, 203–217. https://doi.org/10.1179/174962508X322687
Tronci, L. (2009). Sur les expressions de sentiment en grec ancien. Lingvisticae Investigationes, 32(2), 226–237. https://doi.org/10.1075/li.32.2.06tro
Vives Cuesta, A., & Madrigal Acero, L. (2022). Support-verb constructions in post-classical Greek and sociolinguistics: A diachronic study of εὐχὴν ποιέω as a level-of-speech-marker. In J. M. Baños, M. D. Jiménez López, M. I. Jiménez Martínez, & C. Tur (Eds.), Collocations in theoretical and applied linguistics: From Classical to Romance languages (pp. 305–333). Guillermo Escolar/SEEC.