In grammar, infinitive is the name for certain verb forms that exist in many languages. In the usual (traditional) description of English, the infinitive of a verb is its basic form with or without the particle to: therefore, do and to do, be and to be, and so on are infinitives. As with many linguistic concepts, there is not a single definition of infinitive that applies to all languages. Many Native American languages and some languages in Africa and Aboriginal Australia simply do not have infinitives or verbal nouns. In their place they use finite verb forms used in ordinary clauses or special constructions.
In languages that have infinitives, they generally have most of the following properties :
However, it bears repeating that none of the above is a defining quality of the infinitive; infinitives do not have all these properties in every language, as it is shown below, and other verb forms may have one or more of them. For example, English gerunds and participles have most of these properties as well.
Contents |
English language has three non-finite verbal forms, but by long-standing convention, the term "infinitive" is applied to only one of these. (The other two are the past- and present-participle forms, where the present-participle form is also the gerund form.) In English, a verb's infinitive is its unmarked form, such as be, do, have, or sit, often introduced by the particle to. When this particle is absent, the infinitive is said to be a bare infinitive; when it is present, it is generally considered to be a part of the infinitive, then known as the full infinitive (or to-infinitive), and there is a controversy about whether it should be separated from the main word of the infinitive (see Split infinitive). Nonetheless, modern theories typically do not consider the to-infinitive to be a distinct constituent, instead taking the particle to for operating on an entire verb phrase; so, to buy a car is parsed as to [buy [a car]], not as [to buy] [a car].
The bare infinitive and the full infinitive are mostly in complementary distribution. They are not generally interchangeable, but the distinction does not generally affect the meaning of a sentence; rather, certain contexts call almost exclusively for the bare infinitive, and all other contexts call for the full infinitive.
Huddleston and Pullum's Cambridge Grammar of the English Language (CGEL), published in 2002, does not use the notion of the infinitive, arguing that English uses the same form of the verb, the plain form, in infinitival clauses that it uses in imperative and present-subjunctive clauses.
The bare infinitive is not used in as many contexts as the full infinitive, but some of these are quite common:
The full infinitive (or to-infinitive) is used in a great many different contexts:
When the verb is implied, some dialects will reduce the to-infinitive to simply to: "Do I have to?"
The auxiliary verb do does not have an infinitive — even though do is also a main verb and in that sense is often used in the infinitive. One does not say *I asked to do not have to, but rather, either I asked not to have to or I asked to not have to (but see split infinitive). Similarly, one cannot emphasize an infinitive using do; one cannot say, "I hear him do say it all the time."
Nonetheless, the auxiliary verbs have (used to form the perfect) and be (also used to form the passive voice and continuous aspect) both commonly appear in the infinitive: "It's thought to have been a ceremonial site", or "I want to be doing it already." "I was supposed to have (already) gone" vs "I should have (already) gone."
The modal auxiliary verbs, can, may, shall, will and must are defective in that they do not have infinitives; so, one cannot say, *I want him to can do it, but rather must say, I want him to be able to do it. The periphrases to be able to, to have to and to be going to are generally used in these cases.
There is a specific situation in which the infinitive is used like an "impersonal future tense", replacing "will". This is done through the construction:
Grammatically, this is identical to the instructional "I am to wait outside" construction (above), but does not signify somebody having been issued an instruction; rather, it expresses an intended action, in the same way as "will". This "tense" is used extensively in news reports, eg. –
This "future infinitive" construction is interesting in that it only has a future aspect to it in situations where the speaker is significantly distanced from the event.[3] In cases where the subject of the sentence is not quite as distanced from the speaker, then the same construction takes on a sense of instruction or necessity (as in "he is to wait outside", or "he is to go to hospital").
The same construction can be used in conditional clauses – If you are to go on holiday, then you need to work hard (or, conversely, if you want to...then you are to...).
The impersonality aspect comes from the fact that the emotionless verb to be is used in the place of the more usual modal verbs which would normally connect the speaker to the statement. In this way, statements are given weight (as if some external force, rather than the speaker, is governing events).
Conversely, however, the construction also provides an uncertainty aspect, since it frees the speaker from responsibility on their statement – in the phrase "John will go", for example, the speaker is almost advocating their certainty that John will, in fact, go; meanwhile, "the Prime Minister is to go" simply states the knowledge that the PM's going is in some way foreseen. (If John ends up not going, for example, the "will go" construction is negated, while the PM's "to go" construction would still hold true, since all it expresses is an expectation). In both cases, the knowledge is simply being reported (or pretends to be) from an independent source. In this sense, this impersonal to + verb construction can almost be seen as a fledgeling renarrative mood.
The original Proto-Germanic ending of the infinitive was -an, with verbs derived from other words ending in -jan or -janan.
In German it is -en ("sagen"), with -eln or -ern endings on a few words based on -l or -r roots ("segeln", "ändern"). The use of zu with infinitives is similar to English to, but is less frequent than in English. German infinitives can function as nouns, often expressing abstractions of the action, in which case they are of neuter gender: das Essen means the eating, but also the food.
In Dutch infinitives also end in -en (zeggen — to say), sometimes used with te similar to English to, e.g. "Het is niet moeilijk te begrijpen" → "It is not difficult to understand." The few verbs with stems ending in -a have infinitives in -n (gaan — to go, slaan — to hit). Afrikaans has lost the distinction between the infinitive and present forms of verbs, with the exception of the verbs "wees" (to be), which admits the present form "is", and the verb "hê" (to have), whose present form is "het".
In Scandinavian languages the n has dropped out and the infinitive suffix has been reduced to -e or -a. The infinitives of these languages are inflected for passive voice through the addition of -s or -st to the active form.
The formation of the infinitive in the Romance languages reflects that in their ancestor, Latin, almost all verbs had an infinitive ending with -re (preceded by one of various thematic vowels). For example, in Spanish and Portuguese, infinitives end in -ar, -er, or -ir, while similarly in French they typically end in -re, -er, oir, and -ir. In Romanian the so-called "long infinitives" end in -are, -ere, -ire and they are converted into verbal nouns by articulation (verbs that cannot be converted into the nominal long infinitive are very rare[4]). The "short infinitives" used in verbal contexts (e.g. after an auxiliary verb) have the endings -a,-ea, -e, and -i (basically removing the ending in "-re"). In Romanian, the infinitive is usually replaced by a clause containing the preposition sǎ plus the subjunctive mood. The only verb that is modal in common modern Romanian is the verb a putea, to be able to. But in popular speech, the infinitive after a putea is also increasingly replaced by the subjunctive.
In all Romance languages, infinitives can also be used as nouns.
Latin infinitives challenged several of the generalizations about infinitives. They did inflect for voice (amare, "to love", amari, to be loved) and for aspect (amare, "to love", amavisse, "to have loved"), and allowed for an overt expression of the subject (video Socratem currere, "I see Socrates running").
Romance languages inherited from Latin the possibility of an overt expression of the subject. Moreover, the "inflected infinitive" (or "personal infinitive") found in Portuguese, Galician, and (some varieties of) Sardinian inflects for person and number. These are the only Indo-European languages that allow infinitives to take person and number endings. This helps to make infinitive clauses very common in these languages; for example, the English finite clause in order that you/she/we have... would be translated to Portuguese as para teres/ela ter/termos... (European Portuguese is a null-subject language). The Portuguese personal infinitive has no proper tenses, only aspects (imperfect and perfect), but tenses can be expressed using periphrastic structures. For instance, even though you sing/have sung/are going to sing could be translated to apesar de cantares/teres cantado/ires cantar.
Other Romance languages (including Spanish, Romanian, Catalan, and some Italian dialects) allow uninflected infinitives to combine with overt nominative subjects. For example, Spanish al abrir yo los ojos ("when I opened my eyes") or sin yo saberlo ("without my knowing about it").[5]
In Ancient Greek the infinitive has four tenses (present, future, aorist, perfect) and three voices (active, middle, passive). Unique forms for the middle are found only in the future and aorist; in the present and perfect, middle and passive are the same.
| λύω "I release" |
active | middle | passive |
|---|---|---|---|
| present | λύειν | λύεσθαι | |
| aorist | λῦσαι | λύσασθαι | λυθῆναι |
| future | λύσειν | λύσεσθαι | λυθήσεσθαι |
| perfect | λελυκέναι | λελύσθαι | |
Only the Ancient Greek aorist infinitives active and passive survive in Modern Greek, but their descendants have a totally different function. The Ancient Greek γράψαι "to write" became γράψειν in analogy to the present infinitive γράφειν and then γράψει in Modern Greek and is used only in combination with the auxiliary verb έχω "I have" in the formation of the Present Perfect: έχω γράψει "I have written, lit. I have write(inf.)". When combined with είχα "I had", it yields the Past Perfect είχα γράψει "I had written". Similarly, the Ancient Greek γραφῆναι "to be written" survives as γραφεί (γραφῆ in Katharevousa); thus, έχει γραφεί (ἔχει γραφῆ in Kath.) means "It has been written".
In Pontic Greek, infinitives have a similar function; they only serve for the creation of the Present Perfect Optative: ας είχα γράψ'ναι "I wish I have written". Infinitives are formed this way: active: root of the Future + -ναι; passive: root of the Aorist + -θήν. Examples: εποθανείναι, μαθείναι, κόψ'ναι, ράψ'ναι, χαρίσ'ναι, αγαπέθην, κοιμεθήν.
In Modern Greek, "I want to write" translates θέλω να γράψω (literally, "I want that I write"), opposed to Ancient Greek ἐθέλω γράφειν (literally, "I want to write"). In Modern Greek, the infinitive has changed form and is used mainly in the formation of tenses and not with an article or alone. Instead of the Ancient Greek infinitive "γράφειν", Modern Greek uses the infinitive "γράψει", which does not inflect. The Modern Greek infinitive has only two forms according to voice, "γράψει" for the active voice and "γραφ(τ)εί" for the passive voice.
The infinitive in Russian usually ends in -t’ (ть) preceded by a thematic vowel, or -ti (ти), if not preceeded by one; some verbs have a stem ending in a consonant and change the t to č’, such as *mogt’ → moč’ (*могть → мочь) "can". Some other Balto-Slavic languages have the infinitive typically ending in, for example, -ć (sometimes -c) in Polish, -t’ in Slovak, -t (formerly -ti) in Czech and Latvian (with a handful ending in -s on the latter), -ty (-ти) in Ukrainian, -ць (-ts') in Belarusian. Lithuanian infinitives end in -ti, Slovenian end on -ti or -či, and Croatian on -ti or -ći.
Serbian officially retains infinitives -ti or -ći, but is more flexible than the other Slavs in breaking the infinitive through a clause. The infinitive nevertheless remains the dictionary form. Bulgarian and Macedonian have lost the infinitive altogether (it usually ended in -ти). For that reason, the present first-person singular conjugation is used as the dictionary form in Bulgarian, where as Macedonian uses the third person singular form of the verb.
Hebrew has two infinitives, the infinitive absolute and the infinitive construct. The infinitive construct is used after prepositions and is inflected with pronominal endings to indicate its subject or object: bikhtōbh hassōphēr "when the scribe wrote", ahare lekhtō "after his going". When the infinitive construct is preceded by ל (lə-, li-, lā-) "to", it has a similar meaning as the English to-infinitive, and this is its most frequent use in Modern Hebrew. The infinitive absolute is used for verb focus, as in מות ימות mōth yāmūth (literally "die he will die"; figuratively, "he shall indeed die").[6] This usage is commonplace in the Bible, but in Modern Hebrew it is restricted to high-flown literary works.
Note, however, that the to-infinitive of Hebrew is not the dictionary form; that is the third person singular perfect form.
To form the first infinitive, the strong form of the root (without consonant gradation or epenthetic 'e') is used, and these changes occur:
As such, it is inconvenient for dictionary use, because the imperative would be closer to the root word. Nevertheless, dictionaries use the first infinitive.
There are four other infinitives, which create a noun-, or adverb-like word from the verb. For example, the third infinitive is -ma/-mä, which creates an adjective-like word like "written" from "write": kirjoita- becomes kirjoittama.
The Seri language of northwestern Mexico has infinitival forms which are used in two constructions (with the verb meaning 'want' and with the verb meaning 'be able'). The infinitive is formed by adding a prefix to the stem: either iha- [iʔa-] (plus a vowel change of certain vowel-initial stems) if the complement clause is transitive, or ica- [ika-] (and no vowel change) if the complement clause is intransitive. The infinitive shows agreement in number with the controlling subject. Examples are: icatax ihmiimzo 'I want to go', where icatax is the singular infinitive of the verb 'go' (singular root is -atax), and icalx hamiimcajc 'we want to go', where icalx is the plural infinitive. Examples of the transitive infinitive: ihaho 'to see it/him/her/them' (root -aho), and ihacta 'to look at it/him/her/them' (root -oocta).
In languages without an infinitive, the infinitive is translated either as a that-clause or as a verbal noun. For example, in Literary Arabic the sentence "I want to write a book" is translated as either urīdu an aktuba kitāban (lit. "I want that I write a book", with a verb in the subjunctive mood) or urīdu kitābata kitābin (lit. "I want the writing of a book", with the masdar or verbal noun), and in Demotic Arabic biddi aktob kitāb (subordinate clause with verb in subjunctive).
Even in languages that have infinitives, similar constructions are sometimes necessary where English would allow the infinitive. For example, in French the sentence "I want you to come" translates to Je veux que vous veniez (lit. "I want that you come", with come being in the subjunctive mood). However, "I want to come" is simply Je veux venir, using the infinitive, just as in English. In Russian, sentences such as "I want you to leave" do not use an infinitive. Rather, they use the conjunction чтобы "in order to/so that" with the past tense form (most probably remnant of subjunctive) of the verb: Я хочу, чтобы вы ушли (literally, "I want so that you left").