Basic concepts

The range of meanings coded in English by verbs such as will, want and wish is here comprised under the notion of volition. More precisely, the common denominator is a bivalent predication ‘want (E, Th)’, where Th is the theme, i.e. the entity (object or situation) wanted, and E is the experiencer, i.e. the one who wants Th. E is typically an entity high on the empathy hierarchy. Most commonly, it is the speaker; and this gives rise to specialized grammatical constructions.

Conversion of this notion into linguistic categories is guided by the following criteria:

• Th may be a situation, represented as S^Th, or a thing, represented as N^Th.
• E may be anybody, thus, any of the speech act participants, or may be only the speaker, or may be impersonal.
• E may or may not be (coreferential with) a participant of S^Th.

These conditions place constructions coding a volitive predication at different levels of grammar:

1. S^Th is coded as a clause S^Th, while N^Th is coded as a nominal expression N^Th.
2. If E can be anybody, it is coded as a nominal expression N^E. This functions as the primary argument of a relational expression – a verb or a grammatical operator – coding the want predicate. This yields a desiderative construction.
3. If E is impersonal, it is not coded. The resulting proposition ‘S^Th is wanted’ represents impersonal volition.
4. If E is of necessity the speaker, it need not be coded. The resulting proposition ‘want (I, S^Th)’ is coded as a sentence S^Th with directive illocutionary force.
5. In a variant of #4, E is, again, the speaker, but S^Th is a wish exclaimed without the intention to make anybody present in the speech situation do anything about it; so the utterance lacks directive illocutionary force. Such an utterance is an optative exclamation.

Desiderative predication

In desideration, the predication ‘want (E, Th)’ is the designated situation (instead of operating on it). With such a predication,

1. the desiderative proposition may constitute a sentence of any sentence type or a dependent clause
2. E may be any of the speech act participants; thus, E may shift through the three persons. Consequently, desiderative constructions mostly come out as personal constructions
3. the verb coding want may shift through the TAM categories
4. Th may be a concrete object N^Th (a) or a situation S^Th (b). In the latter case, if want is coded as a verb, this would be a member of the set taking a propositional complement.

   .	a.	How much sugar do you want?
   	b.	Do you want Linda to bring more sugar?

5. E may be (coreferential with) a participant of S^Th. This configuration may be represented by diverse grammatical constructions. Two important types are these:

1. want is coded as a verb, and S^Th is coded in a nonfinite construction, commonly as an infinitival whose empty subject position is under phoric control by E ().

.	Linda wants to have more sugar

The case of N^Th (a) may be construed as a subcase of the present one, viz. as an elliptic version of the structure ‘want (E, (have (E, N^Th))’ (b). This analysis will be abided by in what follows.

A variant of type #a is want as a modal verb. German wollen is a typical representative of this species. Despite its term, this differs from a mood; a modal verb can itself take on different moods.

.	a.	Linda	lebt
German		Linda	lives
		Linda lives

	b.	Linda	will	leben
		Linda	wants	live:INF
		Linda wants to live

2. The configuration is coded by a construction in which the want predicate is an operator on the verb of S^Th. That is, want grammaticalizes to a desiderative modification of the – until then dependent – verb. Such a desiderative form may have various statuses in the morphology:
   • It may be a derivation, as it is in Sanskrit ().
   • It may be verbal enclitic, as it is in Cabecar ().

Although the concept want is coded at the morphological level in type #b, this has nothing to do with mood and modality.¹ Instead, it is structurally on a par with an iterative or semelfactive form of a verb; such derived stems may, again, conjugate for mood or be the operand of different modalities.

Impersonal volition

If E is impersonal, the volitive configuration ‘S^Th is/was/would be wanted/desirable’ results. is a plausible example; its literal meaning is ‘rain is wanted to fall’.

Impersonal volition is by definition a modality. It does not constitute a sentence type, but is compatible with the basic sentence types. In this, it is analogous to debitive and potential modality. These can be satisfactorily construed as monovalent predicates or unary operators ‘necessary (S)’ and ‘possible (S)’, resp. A necessity or a possibility may exist whether or not some E – the speaker in particular – considers them a necessity or a possibility. Thus, deontic modality and epistemic possibility are straightforwardly conceived as impersonal predications. By contrast, an impersonal source of volition is peculiar. If a wish or will exists, it must be somebody's wish or will.

German wollen () is in a paradigm with other modal verbs such as müssen ‘must’ and mögen ‘may’. Some of their uses may be represented by impersonal modal operators on their proposition. In particular, [N muss V] ‘N must V’ can mean ‘it is necessary that N V’; and likewise [N mag V] ‘N may V’ can mean ‘it is possible that N V’. This, however, is never the meaning of German wollen (or English will, for that matter). cannot mean ‘it is wanted that Linda live’.

Commonly, either of two things happens in constructions which could designate impersonal volition:

• While the structure coding the volition is impersonal, the meaning is personal. Yucatec Maya has an operator taak ‘it is wanted’ which takes S^Th as a complement clause ().

.	taak	in	kan-ik
Yucatec	VOL	[SBJ.1.SG	learn-IPFV]
	I want to learn it

Semantically, however, the subject of S^Th represents E. So while this is structurally an impersonal construction, it is semantically a desiderative construction.

• If E in ‘want (E, S^Th)’ actually is impersonal, then S^Th is either necessary or possible. Consequently, the impersonal predication ‘S is wanted’ easily merges with either ‘necessary (S)’ or ‘possible (S)’.
  • The Cabecar [N kia-n-á̱ V] () and likewise the Yucatec [taak V N] () are ambiguous between ‘N wants to V’ and ‘N must V’.
  • The Ancient Greek optative () may code a wish of the speaker, but also an epistemic possibility.

English want illustrates the semantic expansion from ‘want (E, S)’ to ‘necessary (S)’. In addition to its desiderative use, it has acquired a debitive use ().

a expresses the volition of its main subject referent, while #b expresses the obligation under which this puts the subject referent of S^Th. Given that participant E wants S and imposes this will on N, which latter is typically, but not necessarily a participant of S; then E's volitive modality is N's debitive modality. As a result, even if there is no E, the resulting debitive construction may, again, be structurally personal, with its predicate translating as ‘need’.

The net balance is that volitive modality, conceived as impersonal volition, has a shaky position among cognitive and communicative operations.

Directive illocution

The highest level relevant in the analysis of volition is the level of the speech act. Here E of ‘want (E, Th)’ is the speaker. Directive illocution results if the utterance coding ‘want (I, S^Th)’ addresses the interlocutor with the intention that he behave in a way to bring S^Th about. Directive illocution may be marked by a mood, which is the subjunctive in .

There are three kinds of directive sentence depending on the person of the subject of S^Th. A directive sentence

1. with first person subject is hortative (a)
2. with second person subject is precative (#b)
3. with third person subject is jussive (#c)

Case #b is special as it is marked by a dedicated mood in most languages, viz. the imperative. In a language which opposes an imperative () to a second person directive sentence marked otherwise, such as the subjunctive in b, the latter would have precative illocution.

Otherwise, second person directive sentences would simply be imperatives, as generally in English.

Directive illocution constitutes a basic sentence type beside the declarative and the interrogative sentence.

Optative exclamation

In the optative exclamation, E is, again, the speaker. It does not, however, purport to instigate anybody to fulfill S^Th. This is then an illocutionary force which is non-contactual as it does not extend to the hearer. In this, the optative exclamation differs from directive utterances, which do address the hearer.

The non-contactual wish is commonly introduced by an exclamative particle like Spanish ojalá, English would that, Latin utinam () or Ancient Greek eíthe ().

The optative exclamation wishes that things would be otherwise, thus presupposing that S^Th is currently not the case. The optative particle is therefore typically accompanied by a non-indicative mood or modal formative to indicate this. This may be a dedicated mood such as the Ancient Greek optative (). Otherwise, it may be just the subjunctive ().

An optative exclamation may be marked as irreal (counterfactual); i.e., the wish is unfulfillable. If S^Th is in the past, this is necessarily so ().

Optative sentences without a dedicated particle are and a.

This construction does code a wish and can therefore be subsumed under the concept ‘optative’. However, is normally not an exclamation and is therefore sometimes subsumed under an extended concept of ‘precative’. b is a borderline case between a jussive sentence and an optative exclamation.

¹ pace https://en.wikipedia.org/wiki/Desiderative_mood, 10/09/2025