Skip to navigation – Site map

A natural-language semantics approach to infinitival and gerund-participial complementation in English

Patrick Duffley
p. 55-67


Cet article propose une analyse de la complémentation infinitive et gérondive en anglais au moyen d’un cadre théorique fondé sur la sémantique naturelle non formelle. Le postulat de base de l’analyse est que ce sont les signifiés des formes linguistiques et les relations conceptuelles entre ceux-ci qui conditionnent en premier lieu les possibilités expressives des constructions comportant un complément. L’explication proposée de l’emploi des compléments en anglais s’appuie sur trois paramètres : (1) le contenu sémantique des compléments eux-mêmes ; (2) la fonction du complément dans la phrase ; (3) le contenu lexical du verbe de la principale. Dans certains cas, il faut considérer également le contexte d’énonciation pour rendre compte de la production du message. Nous suggérons que l’incapacité d’autres cadres théoriques d’apporter des explications pour certaines données est due au fait qu’ils emploient des catégories sémantiques qui sont ou bien trop abstraites, et donc ne correspondent pas à la sémantique de l’anglais, ou bien trop particulières, ne relevant que de certains emplois des formes en question. Les données choisies pour illustrer l’argument sont prises parmi les verbes aspectuels en anglais.

Top of page

Full text

1. Introduction

1One piece of data concerning the verbs start and begin which has puzzled English grammarians is the fact that whereas start to can be used both when the complement’s event was initiated and when it was not, as in:

(1) All of a sudden it started to rain.
(2) Kestrel started to say something, but decided not to. She would listen until he was finished, but it was not easy to hold her questions, (

2begin + to + infinitive always implies initiation of the infinitive’s event, as in:

(3) The woman giggled and began to say something about a “Miller house” when her husband glared at her and sternly shushed her.

3I hope to demonstrate here that a natural-language semantics approach, allied with the Guillaumian distinction between potential and actual meaning (cf. Hirtle 1975: 5; 2007: 7-8), is capable of handling such facts by means of fairly simple and non-formal meaning-based principles. In the case at hand, three basic parameters are needed in order to account for the semantic and distributional characteristics of infinitival and gerund-participial complementation in English: (1) the semantic content of the complement forms themselves, (2) the lexical content of the matrix verb, (3) the function of the complement with respect to the matrix.

2. The Semantic Content of the Complement Forms

2.1 The Gerund-participle

4The definition of the semantic content of the -ing form has posed a serious challenge to several generations of linguists and grammarians. The two most recent grammars of English, Quirk et al. 1985 and Huddleston & Pullum 2002, avoid the question completely by making no attempt at any unified definition of this form’s meaning at all. When the challenge is taken up, the most common strategy employed is to focus on a prototypical use of the -ing form, most often that found in the progressive, and to define the meaning of the form accordingly, as being “durative aspect” (Kruisinga 1931: 259), “imperfectivizing” (Freed 1979: 72-73), or denoting a mid-point between the beginning and end of an event, as in Hewson & Bubenik’s (1997: 5-6) diagram:

5The most recent cognitive grammar monograph on English complementation follows in the same tradition:

To the -ing gerund, one can attribute precisely the same value that it has in the progressive construction, referring specifically to an activity which is in progress at the moment of time serving as the reference point for the utterance. (…) In She is pruning the roses, the -ing progressive refers to an ongoing activity which is happening at exactly the time of speaking. The activity of pmning takes place over some period of time relative to the moment of speaking. Like the progressive, the -ing gerund takes an internal perspective on the action described by the verb stem, to the exclusion of the initial and final states. (Hamawand 2002: 99)

6There is a basic methodological problem however with the practice of basing the definition of a form’s meaning on just one of its uses and treating its other senses as extensions from this prototype. This practice fails to respect the fact, pointed out repeatedly by Guillaume (cf. Guillaume 1984: 37, 97, 128), that linguistic forms necessarily exist before they are used and that it is their permanent content, detached from any particular context of use and existing as a unitary potential, which explains their utilization to produce the multiplicity of effects observable in discourse. Thus if one takes ongoingness or progressivity as the meaning of the English gerund-participle, one is faced with the problem of how anteriority can be an extension from this meaning in a case such as (4) below:

(4) He admits tripping her up.

7This use is presented by Hamawand as a “second extended meaning… where the complement event temporally precedes the time of utterance expressed by the main verb” (2002: 100). However “anteriority” stands in direct opposition to “ongoingness”, as the latter should always imply simultaneity when put into relation with some other event and consequently never be used to denote something preceding the other verb’s event in time.

8Following Guillaume’s principle that meaning exists as a potential before it is used, the meaning of the gerund-participle proposed here is based on an observation of all of its verbal uses (cf. Duffley 2006: 3-19), with an attempt being made to reconstruct a potential meaning capable, in conjunction with contextual factors, of giving rise to all of these uses. This potential meaning is highly schematic and corresponds, in the case of the English -ing form, to the conceptualization of an event as an interiority in which its implicit subject is contained. This schema can be actualized in various ways. In the progressive construction, the interiority of the event is treated as a property of the subject at a particular moment in time. This produces a representation of the event as divided into an accomplished and a yet to be accomplished portion by the position of the subject within it. The interiority of the event can also be attributed directly to its support as a non-temporalized property of the latter, in which case any effect of incompletion depends on whether the support can possess all (5) or only part (6) of what is involved in the property at the implied reference point:

(5) A man resembling my father walked in.
(stative, no impression of incompletion or progressivity)
(6) A man driving a Mercedes waved at us.
(action, impression of incompletion or progressivity)

9The interiority of an event can also be taken as a whole, in which case it corresponds to a homogenous entity made up of all of the instants contained between its beginning and end limits. This is the effect observed in a use such as (7):

(7) Climbing a mountain is very good for one’s stamina.

10In its use in complement constructions it is this latter actualization of the gerund-participle which is generally found.

2.2 The To plus Infinitive Construction

11The semantic content of the to plus infinitive construction is more complex than that of the gerund-participle, as it is made up of two words—the preposition to and the bare infinitive. The latter is an abstract version of the simple form of the verb and simply situates an event in time, representing event time as containing all of what is involved in the lexical content of the event (cf. Hirtle 1988 and 2007: 86-115 for an analysis of the English simple form in these terms). With an action, this produces a beginning-to-end view of the event, as in He did read my message; with a state, all of the lexical content can be contained in one instant, as in He may be here right now.

12The potential meaning underlying the use of to to introduce an infinitive is the same as that underlying all of its other uses—the notion of a kinetic orientation potentially leading to a point (cf. Guimier 1995-96: 88, who defines it as “a movement of thought from a terminus a quo to a terminus ad quem”). In its use with the infinitive, the terminal point corresponds to the integral actualization of the infinitive’s event, either as an action unfolding from beginning to end or as a state having a full-fledged existence at each instant of its duration. Two basic types of sense effects are observed in discourse, according to whether the movement denoted by to is implied to be realized or not:

(8) Ségolène got to be president.
(9) Ségolène wanted to be president.

13These can be termed “subsequent actualization” and “subsequent potentiality” in order to bring out the constant element of the existence of a before/after relation between the events expressed by the matrix and complement (cf. Duffley 1992: 89).

3. The Function of the Complement with Respect to the Matrix Verb

3.1 The Gerund-participle

14In the uses which will concern us here, the gerund-participle has the function of direct object of the matrix verb, as in:

(10) Many a student fondly remembered listening to her lectures.

15Listening to her lectures is construed in this sentence as ‘that which was fondly remembered’, which accounts for the fact that corresponding passive constructions are possible under certain discourse conditions having to do with the way the passive packages the information conveyed by the utterance (Listening to her lectures was fondly remembered by many a student). The direct object gerund-participle can also be replaced by a pronoun (Many a student fondly remembered that).

3.2 The To plus Infinitive Construction

16The to plus infinitive construction is usually not a direct object, although it can be used in this function, as illustrated by (11):

(11) Everyone considered not to retaliate the best option.

17Here not to retaliate is that which was considered the best option by everyone, as confirmed by the possibility of rephrasing the message using the passive (Not to retaliate was considered the best option by everyone) and the possibility of substitution by a pronoun (Everyone considered that the best option). Generally, however, the to plus infinitive construction following a matrix verb is a goal-or result-specifier expressing that to which the matrix verb’s event leads or is desired to lead. Thus in (12) and (13) below to listen to her lectures does not correspond to that which was got or wanted:

(12) Many a student got to listen to her lectures.
(13) Many a student wanted to listen to her lectures.

18This is confirmed by the absence of passive rephrases and the impossibility of substitution by a pronoun:

(12’) * To listen to her lectures was got by many a student.
(13’) * To listen to her lectures was wanted by many a student.
(12”) * Many a student got that.
(13”) * Many a student wanted that.

19The anaphoric form corresponding to (12) and (13) involves the use of a stranded to (Many a student got to/Many a student wanted to).

20The fact that the to plus infinitive construction is not a direct object but a goal/result-specifier explains why it is found with certain predicates which do not admit direct objects, but which are construable with prepositional phrases introduced by the forward-looking preposition for:

(14a) *We are hoping a raise.
(14b) We are hoping to get a raise.
(14c) We are hoping for a raise.
(15a) *They are striving recognition
(15b) They are striving to be recognized.
(15c) They are striving for recognition.

21The “diastematic” nature of the preposition is visible in these constructions, as it fills a semantic gap between two words which, as can be observed from the examples in (a), cannot be connected directly (cf. Guillaume 1973: 155). The prepositional nature of infinitival to can also be seen in its use to connect the infinitive to a noun, where it is also comparable to other prepositions:

(16a) *a desire recognition
(16b) a desire to be recognized
(16c) a desire for recognition

22Consequently, in its non-direct-object use introducing an infinitive in complement constructions, to defines the semantic relation between the matrix verb’s event and the infinitive’s, representing the latter as the term of a possible movement leading to its actualization.

4. The Lexical Content of the Matrix Verb

23The third explanatory parameter is the lexical meaning of the main verb. With the non-direct-object to plus infinitive construction, the lexical content of the matrix can imply whether the movement denoted by to is understood to be realized or merely possible. Thus in (8) above the notion of ‘achieving a result’ expressed by get implies that the subject of get actually moved to the actualization of be president. On the other hand, in (9) the notion of desire expressed by want leaves the attainment of the goal of the desire merely hypothetical.

24With the direct object gerund-participle, the lexical content of the main verb also plays a crucial role in the production of the overall message. It determines for instance the nature of any temporal relation implied to hold between the matrix verb and complement events. Thus an event which is remembered will be understood to have taken place before the moment of the remembering, as in (10) above. On the other hand, an event which is represented as’enjoyed’will be understood to be enjoyed at the same time as it is being performed:

(17) I enjoyed talking with you yesterday.

25Similarly, an event represented as being considered as a course of action will be implied to be as yet unrealized at the time at which it is being considered:

(18) I am considering buying a new TV.

26And with some verbs, no particular temporal relation will be implied at all:

(19) This sure beats doing it all manually.

27Here the procedure referred to by the gerund-participle phrase is simply represented as being beaten by the procedure referred to by means of the demonstrative.

5. Application to Usage: Aspectual Verbs

28Two pieces of data concerning aspectual verbs which are left unexplained by other approaches will be used here to illustrate the application of the natural-language semantics approach to usage and support its plausibility. The first is the curious fact described in (l)-(3) above that start + to can be used both when the complement’s event was initiated and when it was not, while begin + to always denotes initiation of the infinitive’s event. To this we can add the strange twist that start + -ing is like begin + to in always implying initiation. The second curiosity to be explained has to do with verbs evoking the end of an event. These verbs are all construed with the gerund-participle:

(20) He stopped painting and looked over at her.
(21) He quit painting and looked over at her.
22) He finished painting and looked over at her.
23) He ceased painting and looked over at her.

29However, one of them—cease—can also be construed with the to plus infinitive construction, and is so construed in approximately 90% of its uses in complement constructions:

(24) It ceased to be a problem a while ago.

30The fact to be explained therefore is: why is cease an exception among verbs denoting the termination of an event in being preferentially construed with to plus infinitive?

31No other approach to my knowledge has been able to answer either this question about cease or the curious facts concerning the verb start described at the outset of this paper. The only attempt to do so is found in Freed’s (1979) study of aspectual complementation in English. It represents a paradigm example however of a formal approach that utilizes analytical categories which are referential in nature rather than linguistic-semantic, and so is ultimately unable to provide any plausible explanations of these phenomena. Freed treats aspectual verbs as taking real-world events as their complements rather than propositions or objects. Events are argued to be segmentable into three different stages—onset, nucleus and coda—and aspectual verbs are claimed to be simply a nomenclature for these stages or subparts of them. Based on the fact that one can start to do something but not actually initiate it, while one cannot begin to do something without doing at least a bit of the initial stage of it, Freed (1979: 71, 132) argues that start names the onset of an event, while begin denotes the initial segment of the nucleus.

32The problem with this account is that the majority of the uses of start + to plus infinitive are of the type illustrated in (1) above, in which the infinitive’s event is understood to have been actually initiated. This should not be the sort of message conveyed by start however if its meaning denoted reference to only the onset of the complement’s event. To make matters worse, with a gerund-participle complement, the message conveyed is universally that the complement event was initiated by the subject of start:

(25) *Kestrel started saying something, but decided not to. She would listen until he finished…
(26) He started saying something about how when boys are kids they learn to push back, and I explained that I think women push back too, just differently.

33Clearly, an explanation which treats the semantics of aspectual verbs as a referential nomenclature for segments of an event will not do the job of accounting for the difference between start and begin in complement constructions.

34A natural-language semantics approach can however provide a solution to the puzzle just evoked. The key to the explanation is the lexical distinction between start and begin. Although these two verbs can often be used interchangeably, lexicographers have observed certain differences between them. Hayakawa and Ehrlich point out for instance that start “places more emphasis on the fact of making a beginning, the mere act of setting out”, an observation that is corroborated by Webster’s Dictionary of Synonyms, which characterizes start as suggesting “a setting out from a particular point… often after inaction or waiting”. This calls to mind the use of start with the prepositional particle out (They started out from Whitehorse on February 10), a type of context in which begin cannot be used. Start is also the only member of the pair to be used in a causative sense, as in She started the car, a fact which is pointed out by Freed (1979: 20). In addition, only start occurs in the sense of making a sudden movement due to surprise or fear, as in:

(27) He started in his bed at the sound of the shot.

35These observations show that start is not definable as referring inherently to a segment of an event. Thus while it can be argued that begin does denote “the initial temporal segment of the nucleus of an event” (Freed 1979: 71), it cannot be maintained that start denotes an event’s onset. Rather this verb’s meaning corresponds to the notion of breaking out of a state of rest or inactivity. This makes start an “independent action” (Tobin 1993: 175), while begin refers to the initial part of an activity.

36The lexical distinction between start and begin just described, together with the semantics and function of the complement forms themselves, provide the key to explaining the curious facts of usage that other approaches have been unable to come to grips with. To start with begin, it is quite logical that it should not have any uses of the type illustrated in (2) since it denotes the realization of the initial segment of an event. In the gerund-participle construction, the complement’s event is simply represented as ‘begun’, as in:

(28) They began taking down the tent.

37Representing an event as having its initial segment realized will obviously convey the notion that the event was initiated. With a to plus infinitive complement, as in (29) below, the initiation of the event is conceived in a somewhat different fashion:

(29) They began to take down the tent.

38Here realizing the initial segment of taking down the tent is construed as a movement towards the realization of the whole event, just as driving the first few kilometres of a trip takes one part of the way towards covering the total distance involved in the journey. The preposition to denotes this way of seeing a beginning as a movement, and the bare infinitive following to denotes the full actualization of the event as a prospect towards which the beginning takes the subject. Since one cannot move through the initial segment of an event without initiating the event, begin + to + infinitive always implies that the complement’s event got underway, just as with begin + gerund-participle.

39Moving on now to start, the fact that it denotes a breaking out of a state of rest or inactivity allows one to explain how it can be used in cases such as (2). We have here a situation in which there is a preparatory phase prior to the actual initiation of the saying (in this case, opening one’s mouth in order to speak). This anticipatory phase constitutes a movement towards the actualization of the infinitive’s event, and so the meaning of the to + infinitive structure is perfectly adequate for denoting it. However it is a mistake to conclude from this type of use that start inherently refers to the onset of an event: other uses of this verb show that the beginning of an event can also be construed as a case of breaking out of a state of inactivity. Indeed this is the case in approximately 90% of the uses of the start to construction (cf. Duffley 2006: 106), which correspond to the type illustrated in (1) above. As with the carrying out of the initial segment of the event expressed by begin, in this type of use the breaking out of inactivity denoted by start is construed as a movement towards the realization of the infinitive’s event. The difference with begin lies in the fact that with start the initiating of the event is conceived as a breaking out of a state of inactivity, rather than simply as a going through the initial segment of the event.

40The verb start can also be followed by a direct object, as in:

(30) She started a conversation with the woman sitting next to her.

41The sense expressed by start here corresponds to the notion of causing something to go from a state of non-realization to a state of being in progress. The gerundparticiple after start, as illustrated in (26) above, is also a direct object. Since it denotes that which was started, i.e. caused to go from a state of non-realization to a state of being in progress, it comes as no surprise that the start + -ing construction should always evoke initiation of the gerund-participle’s event.

42It remains now to address the curious behaviour of the verb cease among verbs referring to the end of an event. As with start, the crucial element in the explanation is the lexical meaning of this verb as compared to the other egressives. The most frequent verb in this lexical group is stop, which often poses problems for speakers of Romance languages learning English who tend to use the infinitive after this verb (a French-Canadian hockey coach was once heard encouraging his players Don’t stop to skate!, which must have been rather puzzling for the English speakers among them). The lexical meaning of stop corresponds to the idea of obstructing the passage of something at a certain point. If the thing whose passage is obstructed is an event, the message conveyed will be to evoke the suspension of the latter, as in:

(31) She stopped working on the project.

43Stoppage of an event is obviously compatible with its resumption at some later point. The notion of obstructing something only makes sense with a direct object complement corresponding to that which is/was obstructed, which explains the exclusive use of the gerund-participle after stop.

44The to plus infinitive construction after stop functions as a purpose clause, as illustrated in:

(32) She stopped working on the project to take care of administrative problems in another department.

45In some uses that which is stopped is recoverable from the context and so left implicit as in (33); stop can also be used to denote cessation of some form of movement implicit in the context, as in (34):

(33) He was typing at the computer when I came in. He stopped for a minute to say hello and returned to the task.
(34) We stopped to say hello to Joanna on our way here.

46The verb finish, according to Webster’s (1968: 153), “connotes the completion of the final act in a process of elaboration”. If that which is finished is an event, this will imply the termination of the latter, at least as far as the subject of finish is concerned, as in:

(35) She finished working on the project yesterday.

47Because of its meaning, this verb also makes sense only with a direct object -ing form complement which denotes that which was finished, i.e. that which was completed by applying the final touches to it.

48The verb quit evokes termination as a sharp breaking off or leaving of the agent from the event, which explains why it occurs preferentially with jobs and habits and is often associated with feelings of “disgruntlement and defeat” (cf. Hayakawa & Ehrlich 1994: 438). Unlike finish, it does not evoke the complement event’s completion; compare (35) above with (36) below:

(36) She quit working on the project yesterday.

49The notion of abrupt leaving does however imply unwillingness to return to the situation quit, which accounts for Freed’s impression of a definitive termination (1979: 121). Because of its semantic content, the verb quit can only be construed with a gerund-participle direct object denoting that which is abruptly left and broken off from.

50The verb cease is peculiar among English egressives in being construable with both the gerund-participle and the to plus infinitive construction, a fact for which an explanation will now be proposed. This fact is not the only peculiarity shown by this verb, however, as it also manifests semantic constraints on the type of direct object with which it can be construed that are significant for the problem under consideration here:

(37a) He stopped the car.
(37b) *He ceased the car.
(38a) She finished her essay.
(38b) *She ceased her essay.
(39a) He quit his job.
(39b) *He ceased his job.

51These examples show that, unlike stop, finish and quit, cease implies that its object no longer exists as a result of the cessation, what Hayakawa & Ehrlich (1994: 438) describe as “total extinction”, as illustrated by:

(40) Life magazine has ceased publication.

52Consequently, it will be proposed here that cease denotes the negation of the continuation of the existence of something. This notion is obviously construable with a direct object complement, as in (41) below, where talking corresponds to that which was ceased, i.e. that whose continued existence was negated:

(41) I told him that I’d found the money, relating in detail all I’ve said before. He kept watching me all the time incredulously. (…) But when I’d ceased talking, he said, “Well, Jim, I believe you, I don’t like
(Lancaster-Oslo/Bergen Corpus L15 69)

53Since the existence of an event in time is also conceivable as a movement by which the subject progresses from the past towards the future (cf. Guillaume’s concept of ascending time (1964: 195)), ceasing can also be conceived as not moving on to the next instant of the existence of an event. In this construal, the preposition to is used to denote the possible movement towards the further existence of the event, and cease signifies the negation of this movement. This way of viewing a situation is particularly in harmony with stative events: since a state consists of the repetition of the same situation from one instant to another, it is natural to conceive the termination of a state as the negation of the movement on to its next instant of existence. This explains another curious fact observed by Freed, but left unexplained, which is that cease “is found to occur more comfortably with states than states can occur with many other aspectualizers” (1979: 122).

6. Conclusion

54In conclusion, it is hoped that the preceding discussion, although brief and focussing on certain very specific observations, has demonstrated the explanatory capacity of a natural-language semantics approach to English complementation. This approach has been tested here on data concerning aspectual verbs in English which other frameworks have been unable to account for. The reasons for this incapacity lie in the utilization of semantic categories which do not correspond to the meaningcontent of the forms occurring in the constructions they are meant to explain. This lack of correspondence can be due to a failure to take into account the distinction between meaning as a potential and meaning in use (or, better, the message expressed in a particular use), as in Hamawand’s prototype-based cognitive grammar approach to the -ing form. Or it can be due to the utilization of referential categories defined in abstraction from the linguistic sign, as in Freed’s treatment of the aspectual verbs as a nomenclature for the various segments of real-world events. Whatever the cause, whenever a discrepancy is introduced between linguistic sign and meaning, the result can only be the inability to explain why a particular form is used in certain cases and not in others.

55On the other hand, because the main reason for a form’s use is its meaning, an approach which works with the meaning-content of linguistic items can offer explanations for all of the uses of a form. Of course, in some cases pragmatic factors must also be taken into account in explaining the production of the intended message. However, in order to discern how they operate, one must first have an adequate reconstruction of the meanings of the forms themselves which are used in the utterance. It is hoped that this paper has made a small but concrete contribution to the case in favour of a meaning-based approach to grammatical explanation which works with the semantic potential of the natural-language items occurring in a construction.

Top of page


Duffley, Patrick J. 1992. The English Infinitive. London: Longman.

Duffley, Patrick J. 2006. The English Gerund-participle. A Comparison with the Infinitive. New York: Peter Lang.

Freed, Alice F. 1979. The Semantics of English Aspectual Complementation. Dordrecht: D. Reidel.

Guillaume, Gustave. 1964. Langage et Science du Langage. Paris: Nizet and Quebec: Presses de l’Université Laval.

Guillaume, Gustave. 1973. Leçons de linguistique de Gustave Guillaume 1948-1949. Paris: Klincksieck and Quebec: Presses de l’Université Laval.

Guillaume, Gustave. 1984. Foundations for a Science of Language. Amsterdam: John Benjamins.

Guimier, Claude. 1995-96. A Study in Grammatical Semantics: the Preposition to in English. Sigma 17-18: 83-95.

Hamawand, Zeki. 2002. Atemporal Complement Clauses in English. A Cognitive Grammar Analysis. Munich: Lincom Europa.

Hayakawa, S. I. and Eugene Ehrlich. 1994. Choose the Right Word. A Contemporary Guide to Selecting the Precise Word for Every Situation. New York: Elarper and Collins.

Hewson, John and Vit Bubenik. 1997. Tense and Aspect in Indo-European Languages. Amsterdam: John Benjamins.

Hirtle, Walter H. 1975. Time, Aspect and the Verb. Quebec: Presses de l’Université Laval.

Hirtle, Walter H. 1988. Events, Time and the Simple Form. Revue québécoise de linguistique 17: 85-106.

Hirtle, Walter H. 2007. Lessons on the English Verb. Montreal & Kingston: McGill-Queen’s University Press.

Huddleston, Rodney and Geoffrey K. Pullum 2002. Cambridge Grammar of the English Language. Cambridge: Cambridge University Press.

Kruisinga, Etsko. 1931. A Handbook of Present-Day English, Part II. English Accidence and Syntax, vol. 1, 5th ed. Groningen: P. Noordhoff.

Lancaster-Oslo/Bergen Corpus of British English. 1978. Oslo: Department of English, University of Oslo.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman.

Tobin, Yishai. 1993. Aspect in the English Verb. London: Longman.

Webster’s New Dictionary of Synonyms. 1968. Springfield, MA: G. & C. Merriam Co.

Top of page


Bibliographical reference

Patrick Duffley, « A natural-language semantics approach to infinitival and gerund-participial complementation in English », Anglophonia/Sigma, 11 (22) | 2007, 55-67.

Electronic reference

Patrick Duffley, « A natural-language semantics approach to infinitival and gerund-participial complementation in English », Anglophonia/Sigma [Online], 11 (22) | 2007, Online since 13 December 2016, connection on 25 May 2018. URL : ; DOI : 10.4000/anglophonia.775

Top of page

About the author

Patrick Duffley

Université Laval (Quebec)

Top of page


Licence Creative Commons
Anglophonia – French Journal of English Linguistics est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
  • Logo Presses universitaires du Midi
  • OpenEdition Journals