Pinkster, Harm (1942-) [1990], Latin Syntax and Semantics [info], xii, 320 p.: ill.; 24 cm. [word count] [Pinkster].
Previous Section

Next Section

10 Sentence type, illocutionary force and mood

In chapters 1–9 we have dealt with the internal structure of sentences and parts of sentences, e.g. the relation between the predicate and the arguments and satellites, the types of argument and satellite, the formal means to mark the relation between such constituents. This chapter deals with properties of the sentence as a whole (e.g. the distinction between declarative and interrogative sentences) and the way in which the various sentence types are used in specific communicative situations.

In this chapter I first discuss the distinction between sentence type, illocutionary force and mood ( crosssection 10.1.). Then sentence type and illocutionary force are treated separately. In discussing the sentence types I also devote attention to the illocutionary forces (these are, thus, discussed twice) and the moods in the various sentences types ( crosssection 10.2). crosssection 10.3. concerns the relation between the various uses of the moods, e.g. the occurrence of the individual moods in main sentences and subordinate clauses.

10.1. Sentence type, illocutionary force, mood: definitions

(a) Sentence type

Examples (1) and (2) contain the same predication. In both instances the predicate facere is construed with a tu (Subject Agent) and a hoc (Object Patient). This predication can be represented as follows:

facere (tu)AgSubj (hoc)PatObj

(1) and (2) differ, however, in `sentence type' (or, in the terms of crosssection 7.1.2, in `modality').

(1) tu hoc facies (`You will do this')

(2) tune hoc facies? (`Will you do this?')

(1) is a declarative sentence, (2) is an interrogative sentence. In English this difference appears from a difference in word order and intonation. In the Latin

-- 190 --

example the difference becomes clear from the presence of the enclitic question particle -ne. In (3) we find the same predication, this time in an imperative sentence with a specific verb form. I return to the criteria used to distinguish these three sentence types in crosssection 10.2.1.

(3) hoc fac (tu) (`(you) do this') [1]

(b) Illocutionary force

A speaker or writer may use the same sentence with different intentions. Example (1), for instance, is first of all an assertion of the speaker/writer with regard to an action that will be realized in the future. The sentence can, however, also be intended as an order. The hearer is supposed to interpret the assertion with regard to his future activities as an order (so-called futurum pro imperativo, see crosssection 10.2.2.2). In this case, we say that (1) has a directive [2] illocutionary force. In other words, the assertion indirectly functions as an order: in this interpretation, (1) is equivalent to (3).

There is a certain relation between sentence type and illocutionary force. Certain illocutionary forces occur more often with certain sentence types than with others. A declarative sentence will usually be interpreted as the assertion of a fact of which the speaker more or less guarantees the validity (or pretends to)-so with an assertive illocutionary force-rather than as a request for information. I return to the ways in which the illocutionary force of a sentence can be made explicit in crosssection 10.2.2.1).

(c) Mood

In a more general sense, `mood' is sometimes used for the phenomena described under (a) and (b), but it is usually used in its more specific sense for the formal characteristics of finite verb forms (indicative, subjunctive, imperative). I use `mood' only in the latter sense. It is a `morphosemantic' verbal category that indicates the speaker's attitude with regard to the content of the predication as a whole (see Sz. 325). In this way, mood may be compared to tense (see chapter 11). Both are characteristics of the predication that are expressed in the finite verb form. There is a certain relation between mood and sentence type. The imperative mood, for instance, practically only occurs in imperative sentences. [3] See table 10.1. Table 10.1.
Mood
↓Sentence typeImperativeSubjunctiveIndicative
Declarative-++
Imperative++-
Interrogative-++

-- 191 --

It is obvious that there is a relation, albeit to a limited degree, between mood and sentence type. For the sentence type also may be viewed as a means to characterize a predication formally in accordance with the speaker's point of view. [4]

(d) The current approach to sentence type, illocutionary force and mood

Most Latin linguists do not distinguish sentence type, illocutionary force and mood. [5] In most Latin grammars what I call `illocutionary force' tends to be described as specific uses of the moods; e.g. the use of example (1) as an indirect order, where the grammars speak of `indicativus futuri proimperativo' (see Sz. 310–1; 326–7). Thus, the occurrence of the subjunctive in sentences with various illocutionary forces leads to a distinction between the use of the subjunctive in orders, exhortations, etc., as if there were different subjunctives. I return to this in crosssection 10.3.

10.2. Sentence type and illocutionary force

In this section I discuss the various sentence types to be distinguished, the illocutionary forces with which they may be used and the use of the moods in sentences of the various types.

10.2.1 Sentence types; criteria

Three sentence types are commonly distinguished in Latin: declarative, interrogative and imperative sentences. [6] Moreover, like English, Latin has exclamative sentences. As has already been pointed out above (p. 189), declarative and interrogative sentences differ from one another in the addibility or presence with the latter of question particles (-ne; num; nonne) or, in the case of `wh-questions', in the presence of interrogative words such as quis?, ubi?, qualis?. Besides, in spoken Latin intonation, and possibly also word order, will have played an important role. If, however, such formal characteristics are lacking, the sentence type cannot be determined. In such cases, the context is the only means to determine whether we are dealing with a statement or a question, e.g. (4)–(5):

(4) heus tu, Rufio … cave sis mentiaris: Clodius insidias fecit Miloni? :: fecit (`Look here, Rufio … mind you don't tell lies! Did Clodius plot against Milo? He did', Cic. Mil. 60)

(5) (`In Sparta boys are hardened e.g. with the whip, without anyone uttering a sound') quid ergo? hoc pueri possunt, viri non poterunt? (`So? Children can suffer this and men will not be able to?', Cic. Tusc. 2.34)

In (4) the interrogative nature of the sentence Clodius … Miloni appears from the fact that it is followed by an answer. Instances such as (5) are more difficult,

-- 192 --

since there is no answer. Here the preceding quid ergo indicates that a request for information or a rhetorical question will follow. Declarative and interrogative sentences also differ in terms of the types of expression that may occur in them. In interrogative sentences, for instance, we do not find reinforcing expressions such as edepol, hercle and ecastor. [7]

Declarative and interrogative sentences can be distinguished from imperative sentences by their different type of negation. The former sentence types are negated by non, the latter by ne. This is already observed by Quintilian (Inst. 1.5.50). [8] The grammars mention exceptional instances with non in prohibitive sentences, offering as an explanation that these are cases of so-called `Begriffsnegation', constituent negation. A very small number of instances is explained by assuming that non is seen as a stronger variant of ne (the reverse is not found). All instances mentioned, however, involve the explicit negation by non of a presupposition contained in context or situation. [modified 01-09: Examples are (6a) and (6b):

(6a) (Make sure that nurses of young children use the language correctly) non adsuescat ergo, ne dum infans quidem est, sermoni qui dediscendus sit (`So do not let him get used, not even when he is still a child, to a language that must be unlearned', Quint. Inst. 1.1.5)

(6b) quoniam omnia commoda nostra … salutem denique legibus obtinemus, a legibus non recedamus. (`since it is the laws that give us all our advantages ... and our security, let us abide by the laws', Cic. Clu. 155) ]

In this way, non ergo or non igitur is sometimes used to indicate that on the basis of a known fact it should not be presupposed that …. [9]

Besides negation, intonation presumably played a part as well [10] and possibly also a difference in word order (see crosssection 9.2.3. on p. 167). In a text it is in most cases completely clear whether a sentence is imperative or declarative. But there are also instances in which one may doubt, particularly in sentences with a predicate in the subjunctive. It is, for instance, sometimes difficult to decide whether such a predicate is–in traditional terms–a `potential subjunctive' (declarative sentence, negation non) or a `jussive subjunctive' (imperative sentence, negation ne). Examples are expressions such as maneam, opinor (`I think I'll stay') and abeam puto (`I'll go, methinks'), which are often translated `jussively', but might just as well be `potential' (also considering the meaning of opinor and puto).

Exclamative sentences undoubtedly differed in intonation from the sentence types mentioned above. In exclamative sentences we find expressions such as qui, qualis, ut (formally identical with question words) and sometimes the enclitic particle -ne, formally identical with the question particle -ne. In reality, a number of subtypes may be distinguished (see crosssection 10.2.1.4).

I now discuss each of the four sentence types distinguished here, devoting attention to the illocutionary force with which each sentence type occurs (as appears from, among other things, the presence of certain particles) and the mood.

-- 193 --

10.2.1.1 Declarative sentences

(a) Illocutionary force, particles, etc.

Declarative sentences can be used in certain communicative situations with an illocutionary force that is different from that which is normally associated with this sentence type. [11] Examples are (7) and (8):

(7) itis, paratis arma quam primum viri (`Come on, men, get ready for the battle as soon as possible', trag. inc. frg. 34–from Sz. 327)

(8) tu tamen … ut adhuc fecisti, nos consiliis iuvabis (`You must help me with your advice, as you have done up to now', Cic. Att. 10.2.2)

For more examples see K. -St.I.144; Sz. 310–11. In declarative sentences we may, however, also find modal particles that merely modify, i.e. downtone or amplify, the `assertive' character. English examples are really and just in (9) and (10):

(9) This is really too crazy

(10) This is just too crazy

Conversely, question particles (e.g. perchance?) and directive particles as a rule do not occur in declarative sentences. For Latin little systematic research has been done on this point. Amplifying particles in declarative sentences might be profecto, equidem and possibly other of the so-called Modalitätsadverbien in K. -St. I. 795 ff. (see above p. 32). See table 10.2 (p. 204) for some Latin particles that seem to occur exclusively or predominantly in directive sentences (sis, sodes). Further research is desirable. [12]

(b) The use of the moods

Declarative sentences may have a predicate in the indicative or in the subjunctive. [13] That in spite of the difference in mood both types are declarative appears, among other things, from the negation (non) and from the possible presence of modal particles, as mentioned above, and of adverbs that express the judgment of the speaker (i.e. in the function Disjunct).

The difference in mood reflects a difference in meaning. With the indicative the speaker indicates that he regards (or pretends to regard) the statement as true, as factive; with the subjunctive the speaker indicates that he is not responsible for the `factiveness' of the statement, but, for instance, that he regards it as possible. This use of the subjunctive is called `potential'. I limit myself to giving some examples of sentences with the subjunctive, as instances with the indicative are sufficiently clear in themselves.

(11) vix verisimile fortasse videatur (`It may perhaps seem hardly likely to you', Cic. Fam. 7.2.3)

(12) hoc tibi … primum persuadeas velim (`I would like you first of all to be convinced of this', Cic. de Orat. 3.83)

-- 194 --

(13) cuius ego iudicium, pace tua dixerim, longe antepono tuo (`Whose judgement I, by your leave, put by far before yours', Cic. Tusc. 5.12)

(14) sed fortasse dixerit quispiam tibi propter opes… tuam tolerabiliorem senectutem videri (`But somebody might perhaps say that thanks to your riches your old age seems more bearable', Cic. Sen. 8)

(15) at eius avunculum vix intellegeres id agere, cum ageret tamen, Africanum (`But you could hardly have thought that his uncle, Africanus, occupied himself with this', Cic. de Orat. 3.87)

(16) si gladium quis apud te sana mente deposuerit, repetat insaniens, reddere peccatum sit (`If somebody in full possession of his faculties has left with you a sword, and asks it back when he is out of his mind, then giving it back would be a mistake', Cic. Off. 3.95)

Declarative sentences in the subjunctive express the statement in a more careful way. This appears from the frequent presence of expressions that limit the validity of the statement (e.g. fortasse) and of conditional clauses, e.g. the si-clause in (16).

In (15) we find the imperfect subjunctive. Strictly speaking, the form intellegeres locates the action in the past (as does the indicative, see crosssection 11.2.2.1), as a past, and therefore more remote, possibility. Sentences with the present subjunctive and sentences with the imperfect subjunctive differ in that in the latter case the action is interpreted as less obvious or even purely hypothetical. [14] In many such instances we also find a si-clause. We then speak of an `irrealis'. An example is (17):

(17) cuperem equidem utrumque, si posset (`I should like both, if it were possible', Cic. Tusc. 1.23)

We will encounter this, in a sense `atemporal', use of the imperfect subjunctive in other sentence types as well. [15] In a similar way, the pluperfect subjunctive may be used `atemporally'. 15a

In (14) we find the form dixerit. The possible difference in meaning between the perfect subjunctive and the much less frequent present subjunctive is a matter of dispute. According to one view, there is a difference in terms of `aspect' (see crosssection 11.2.1.2 on p. 233). This would mean that dixerit expresses `completeness' or-from a slightly different point of view-`anteriority' with regard to the moment of utterance. There is no evidence for this. Another explanation is based on the observation that in Plautus and Terence the perfect form is rare (see Sz. 333), [16] but relatively frequent in Cicero. Many linguists think that Cicero has consciously extended the use of this form, in order to create a Latin counterpart of the Greek ειποι αν τις as a polite phrase in his dialogues. But this explanation does not apply very well to dixerim in (13), where a Greek counterpart is lacking. [17] A third view maintains that the perfect form is a milder, less explicit variant of the present form; one of the

-- 195 --

arguments adduced is the fact that in Plautus underlings use this form when addressing their superiors. [18] The material does not allow this conclusion. One may wonder whether one explanation can be found that covers all the various uses of the perfect form. We have already seen that there are differences in frequency of use between Plautus and Cicero. The phenomenon especially occurs with a limited number of verbs, among which there are non-productive forms such as ausim (`I would dare'), faxim (`I would do'). In the second person the perfect is not (or hardly ever) found. Also, the present and perfect forms are not interchangeable in all constructions: it is pace tua dixerim, and not dicam. [19]

10.2.1.2 Interrogative sentences

(a) Illocutionary force, particles, etc.

Commonly a distinction is made between `sentence questions' (or `yes/no questions' or `neutral questions') and `word questions' (or `wh-questions'). Word questions are questions that contain an interrogative pronoun, adjective or adverb (quis, uter, qualis, ubi). Such sentences enquire about the identity of a certain constituent. Sentence questions aim at the confirmation or denial of the sentence as a whole. These can be introduced by interrogative particles (-ne, num, nonne, sometimes an). [20] The answer to a sentence question can be, among other things, a so-called `sentence adverbial' (e.g. certe, fortasse) (see p. 32). [21]

The interrogative sentence type is generally used as a request for information (with `information-requesting' illocutionary force). From the fact that interrogative sentences inherently presuppose a lack of information on the part of the speaker it follows that expressions that imply the possession of information cannot occur in questions (e.g. sine dubio, haud dubio, minime, etc.). Certe only occurs in formulas such as certene vides? (`Surely you see?'). Fortasse (`perhaps') occurs in interrogative sentences only on the noun phrase level. For edepol etc. see above p. 192. In interrogative sentences one does, of course, expect to find interrogative particles (besides the particles -ne, num, nonne mentioned above). Here, too, further research is required.

Interrogative sentences can, however, also be used with other illocutionary forces, e.g. with an assertive illocutionary force (so-called rhetorical question). In such instances we do expect to find expressions that show that the addressee is supposed to know the answer, e.g. enim in (18):

(18) quo enim se, repulsos ab Romanis, ituros (`For to whom would they turn, after having been rejected by the Romans', Liv. 34.11.6)

For examples of `rhetorical questions' see K. -St. I.178. [22]

Furthermore, we can distinguish `echo questions'. Such questions follow a declarative or imperative sentence. Examples are (19)–(20) (see also (27)–(28) below):

-- 196 --


(19) audi :: ego audiam (`Listen :: should I (listen)?', Ter. An. 894)

(20) numquam auferes hinc aurum :: atque iam dabis :: dabo? (`You will never take the gold away from here :: and yet you shall give it to me :: shall I?', Pl. Bac. 824–5)

The illocutionary force of echo questions is often negative-assertive: `I will not listen' and `I will not give it to you', respectively.

Interrogative sentences may also have a directive illocutionary force (see Sz. 467), [23] particularly questions in the second person introduced by quin. An example is (21):

(21) quin quiescis? (`Will you calm down?' Pl. Capt. 636)

Questions with non also often have a directive illocutionary force (those with nonne only rarely).

(b) The use of the moods

As in declarative sentences, also in interrogative sentences we find both indicative and subjunctive forms, again with a difference in meaning. Examples are (22)–(26). I limit myself to one example of the indicative.

(22) cui dono lepidum novum libellum (`To whom do I give this nice new booklet', Catul. 1.1)

(23) o me perditum, o afflictum! quid nunc? rogem te, ut venias, mulierem aegram et corpore et animo confectam? non rogem? (`Oh, I am finished, beaten! Now, what can I do? should I ask you to come, you, a sick woman, afflicted in mind and body? Should I not ask you?', Cic. Fam. 14.4.3)

(24) quid ego facerem? (`What should I have done?', Pl. Mer. 633)

(25) putaresne umquam accidere posse ut mihi verba deessent? (`Could you have thought that it would ever happen that I would be lost for words?', Cic. Fam. 2.11.1)

(26) quis enim non fateatur (`Who would not recognize …', Cic. Brut. 279)

In (22) the poet simply asks himself to whom he shall dedicate the book. Hence the indicative. The sentence might also be interpreted as an echo question. [24] In (23), however, we find a so-called deliberative subjunctive, in reality - still in the traditional terminology-a potential subjunctive in an interrogative sentence. The special feature of these questions is that they are put to the questioner himself. Examples (27)–(28) are usually treated in close connection with (23):

(27) ne fle :: egone illum non fleam (`Don't cry! :: Should I not cry for him?', Pl. Capt. 139)

-- 197 --

(28) iurandumst tibi :: quid iurem? :: quod iubebo (`You must swear :: What must I swear? :: What I will tell you', Pl. Rud. 1334)

These echo questions can be regarded as the repetition of a preceding order in the form of a question; thus, for example, K.-St. (I. 181) treat them as instances of the jussive subjunctive in the form of a question. [25] But we find only the negation non (and even this only rarely), so that it is unattractive to interpret these sentences as `interrogative orders' (for examples see K.-St. I. 181; II. 508–11). (24) is also a question put to the questioner himself (`deliberative subjunctive'), but in this case with reference to the past.

As for mood and tense, we see in interrogative sentences, irrespective of their illocutionary force, the same differences between present indicative, present subjunctive and imperfect subjunctive as with declarative sentences: factive, possible, counterfactual. The difference between the imperfect subjunctive and the present subjunctive is neatly illustrated by the following two sentences, the first a question of Cicero's, the second the answer of his interlocutor Brutus:

(29) quid tu, Brute, posses, si te … contio reliquisset (`What could you do, B., if the audience had abandoned you?', Cic. Brut. 192)) [26]

(30) tamen si a corona relictus sim, non queam dicere (`Nevertheless, if I should be abandoned by the public, then I could not speak', ibid.)

Cicero is very cautious and merely formulates an imaginary hypothesis. Brutus, on the other hand, leaves open the possibility.

10.2.1.3 Imperative sentences

(a) Illocutionary force, particles, etc.

Imperative sentences are usually understood to be sentences that express an order of the speaker/writer to an addressee (i.e. a `second party') In this chapter, however, the imperative sentence as a sentence type has been defined on the basis of the negation ne. A common feature of imperative sentences is the fact that all kinds of expressions with regard to the validity of the predication cannot occur (e.g. fortasse, iure). Also, expletives such as ecastor, edepol, mecastor and hercle are not, or rarely, found. [27] There are, however, also differences between the various subtypes. Imperative sentences not only comprise sentences with a strictly directive illocutionary force, but also sentences that express a wish (including imprecations and curses; see chapter 2, n. 16), a permission or concession, etc. [28] I now discuss three subtypes: imperative sentences with a directive, optative and concessive illocutionary force.

(i) Imperative sentences with a directive illocutionary force are-besides the restrictions mentioned above-characterized by restrictions on the type of predication, see p. 17 ff. One such restriction is that of `controllability' (see

-- 198 --

crosssection 2.4.). [29] Another may be found in the fact that it is not very well possible to give orders for a situation that is anterior to the moment of utterance. An exception might be constituted by sentences such as (31), if, as is often done, we regard them as imperative sentences: [30]

(31) quid facere debuisti? :: quod superaret pecuniae rettulisses … solvisses … frumentum … ne emisses (`What should you have done? You should have returned the rest of the money … you should have paid … you should not have bought the corn', Cic. Ver. 3.195)

For examples see M.-St. I.187. In imperative sentences with an unequivocally directive illocutionary force we find words such as age, modo, quin, sis, amabo and the enclitic -dum. [31]

I give a number of examples of what I call–for the sake of brevity–`true orders'. The grammars sometimes further distinguish between first person subjunctive forms and second and third person forms. I do not make this distinction, because such subtypes seem to differ only in person. It is, of course, true that for the second person both the subjunctive and the imperative were available; the choice between these will be discussed below, in section (b) on p. 199. For more examples see K.–St. (subjunctive: I. 180; I. 185–7; for prohibitions I.187–9; imperative: I.195–9).

(32) age sis eamus, nos curemus (`Come on, let us go and take care of ourselves', Pl. Poen. 1422

(33) <suum> quisque igitur noscat ingenium (`Everybody should know his own talents', Cic. Off. 1.114)

(34) eae (litterae) te ne moverint (`Do not let this letter upset you', Cic. Att. 16.1.6)

(35) sis fortis quoad rei natura patiatur (`Be strong, as far as the situation allows it', Cic. Q. fr. 1.3.10)

(36) bene valete et vivite, bene, quaeso, inter vos dicatis mi med absenti tamen (`Good luck and live happily, please speak well about me in your conversations, even though I am no longer there', Pl. Mil. 1340–1)

(37) ne sis plora (`Please do not cry, please', Pl. Per. 656)

(38) i, sequere illos, ne morere (`Come on, follow them, do not dawdle', Pl. Mil. 1361)

(39) proin tu ne quo abeas longius ab aedibus (`Mind you, do not go too far away from home', Pl. Men. 327)

(40) ne sis ferro parseris (`Please do not spare your sword', Pl. Per. 572)

(41) ne vos quidem, iudices, mortem timueritis (`You, judges, must not fear death', Cic. Tusc. 1.98)

-- 199 --

(42) quin tu hoc crimen aut obice … aut iacere noli (`Then bring the charge or drop it', Cic. Q. Rosc. 25)

(43) noli sis tu illi advorsari (`Please do not resist him', Pl. Cas. 205) 31a

(ii) Imperative sentences with an optative illocutionary force (wishes) resemble true orders in some aspects, but differ in terms of controllability and anteriority. There are no restrictions on these points. Of the particles mentioned above most do not occur with wishes.

(44) valeant, inquit, valeant cives mei; sint incolumes; sint florentes; sint beati (`May my fellow-citizens be well, he said, may they be unharmed, may they prosper, may they be happy' Cic. Mil. 93)

(45) utinam ego, inquit, tertius vobis amicus adscriberer (`If only I could be added to you as a third friend', Cic. Tusc. 5.63)

(46) utinam vere … auguraverim (`May I have predicted this correctly', Cic. Rep. 4.8)

(47) tu vel suda vel peri algu, vel tu aegrota vel vale (`May you sweat or freeze to death, or be ill or healthy', Pl. Rud. 582)

For examples see K.–St. I. 182–5.

Wishes need not be addressed to a certain person of whom activities are desired in order to create the situation wished by the speaker. The necessity to interpret such sentences as optative often [32] appears from the presence of utinam, as in (45). 32a

(iii) Imperative sentences with a concessive illocutionary force (concessions) are not subject to restrictions on controllability and anteriority either. Examples are (48)–(50):

(48) haec si vobis non probamus, sint falsa sane, invidiosa certe non sunt (`If we cannot convince you of these things, let them be incorrect; they certainly are not reprehensible', Cic. Luc. 105)

(49) nemo is, inquies, umquam fuit. ne fuerit (`No one like that, you will say, has ever existed. Let this be so', Cic. Orat. 101) [33]

(50) mea quidem hercle causa vidua vivito (`As far as I am concerned, you may damned well live like a divorcee', Pl. Men. 727)

For examples see K.–St. I. 189–91; 199. The necessity of the concessive interpretation often appears from the presence of sane (e.g. (48)). Note that in (49) the negation is ne. [34]

(b) The use of the moods

As appears from the examples, the subjunctive occurs in all subtypes of

-- 200 --

imperative sentences; consequently, these subtypes cannot be distinguished on the basis of the subjunctive alone. Furthermore, Latin has imperative forms and periphrastic expressions such as fac (venias) and noli (venire). [35] There is a great deal of literature on the differences between the various means available to express an order or a prohibition. An additional difficulty in establishing these differences is caused by the existence of diachronic differences between authors. I discuss a number of cases. 35a

(i) The difference between the present subjunctive and the present imperative in expressing an order (facias, fac) is described, from antiquity onwards, as a difference between an advice and an order (e.g. Charisius 228 K) and sometimes even now as a difference in politeness, or mildness, with facias as the milder form. [36] Others think that there is no difference. [37] They point out, for example, that instances such as (51) cannot very well be regarded as examples of mildness. They also call attention to metrical factors.

(51) easque in maximam malam crucem (`May you go to the worst possible hell', Pl. Cas. 611)

(51) is, however, an imprecation rather than an order, and cannot, therefore, very well be adduced as a counter-argument. Imprecations at any rate allow a great deal of uncontrollability (e.g. (49)). The occurrence of the imperative and the subjunctive in one textual unit might prove that there is no difference in meaning. The instances are, however, limited in number. [38] For a good evaluation of the two expressions one should at any rate realize that the subjunctive is relatively rare. [39] The `stronger' character of the imperative might also be deduced from the fact that we often find age, agedum, sis, quin with the imperative, but not very often with the subjunctive (but statistically this is not surprising). For other expressions such as quaeso, proin there is no difference. Finally, idiomatic preferences also seem to play a role–unless there is a more profound reason: in Plautus we find vide rather than videas, fac/facito (ut) rather than facias (ut). Clearly, further research is required.

39a

(ii) The form facito (so-called future imperative) is usually understood as an order for the more distant future. [40] In Plautus, Terence and Cato the form in -to is practically only used for the second person. For the third person the -to- form is often found in juridical texts. It then refers to a general precept, rather than to something that must be executed immediately. See example (52):

(52) si pater filium ter venumduit, filius a patre liber esto (`If the father sells the son thrice, then the son must be taken from the father', Lex XII 4.2)

Furthermore, this form is used in more or less idiomatic expressions such as facito ut memineris (`make sure you remember', e.g. Pl. Aul. 257), where, of course, an instruction for the more distant future is involved. [41] 41a

-- 201 --

(iii) For prohibitions in Latin four means were available: ne fac (example (37)); ne facias (examples (38)–(39)); ne feceris (examples (40)–(41)); noli facere (examples (42)–(43)). The frequency of these expressions varies. The expression ne fac practically does not occur outside poetry after Plautus and Terence, where it is frequent (Sz. 340). From antiquity onwards it is described as a direct prohibition, in opposition to ne facias, which is described as an advice (Charisius 228 K, see also above p. 200). 41b It seems as if the expressions of the type ne fac are used with a limited number of verbs, viz. those indicating a state of mind, but this is presumably a coincidence. According to some scholars, in Plautus expressions of this type always mean `stop … -ing', i.e. they constitute an order to interrupt an activity that is already going on. But there are exceptions. [42] Further research is required in order to determine whether this interpretation (`stop … -ing') is excluded with the other expressions.

Of the pair ne facias/ne feceris the former is the less frequent. [43] It is often regarded as the milder, more advisory variant, while ne feceris is regarded as a categoric order. Incidentally, already in Plautus this semantic difference is difficult to prove owing to the small number of instances. Ne fac is the normal expression there. In Cicero both ne fac and ne facias are hardly ever found, so that it is useless to call ne feceris a stronger `variant'. According to many linguists, there was at least in a prehistoric stage an aspectual difference between the two forms (see also crosssection 11.2.1.2 on p. 233). The relatively high frequency of ne feceris in Cicero is sometimes also ascribed to Greek influence. See also the explanation of dixerit on p. 194. [44]

Noli facere, finally, is considered the most polite form of prohibition. Sz. (336–7) points at the occurrence of expressions like quaeso, sis, edepol in this type of order in Plautus, but in reality these expressions also occur with other types of prohibition. Further research is required. [45] Comparatively speaking, the noli-variant is more often used when the addressee is a judge than the ne feceris-variant. It is also remarkable that Cicero uses ne feceris more often in his letters to Atticus and Quintus than the 'Letters to his friends' (Fam.), which are generally speaking more formal (see Sz. 337). [46] 46a

10.2.1.4 Exclamative sentences

In Latin there are two types of exclamative sentence:

(a) sentences introduced by words like qui, qualis, quantus, quam, quot, ut, etc., usually followed by a finite verb form or only a noun phrase in the nominative;

(b) sentences consisting of a noun phrase in the accusative + in some cases an infinitive.

Examples of these two types are (53)–(56) and (57)–(60), respectively:

(53) quotiens tibi iam extorta est ista sica de manibus! (`How often this dagger has been wrested from your hands', Cic. Catil. 1.16)

-- 202 --

(54) quas ego hic turbas dabo! (`What a mess I will make here', Pl. Bac. 357)

(55) en quibus familiis quam foedis, quam contaminatis, quam turpibus dedatis hanc familiam, iudices! (`See now, judges, to what families, how foul, how polluted, how degraded, you would surrender this family', Cic. Scaur. 13)

(56) qui comitatus in inquirendo! Comitatum dico. Immo vero quantus exercitus! (`What a crowd were involved in the enquiry! "Crowd", I say. Nay, what an army!', Cic. Flac. 13)

(57) O mortalem beatum! cui certo scio ludum numquam defuisse (`Oh happy mortal! I know that he has never been at a loss for jokes', Cic. Div. 2.30)

(58) me caecum, qui haec ante non viderim (`I am blind not to have seen this beforehand', Cic. Att. 10.10.1)

(59) o tempora, o mores (`What times, what customs', Cic. Ver. 4.56)

(60) huncine hominem tantis delectatum esse nugis, ut … (`That this man derived enjoyment from such trifles that …', Cic. Div. 2.30)

At first sight, the first type of exclamative sentence resembles word questions, owing to the occurrence of qu-words (but see n. 22). But in exclamative sentences with quam + adjective (e.g. (55)) we only find gradable adjectives with an evaluative meaning (foedus (`foul'), turpis (`ugly'), etc.), a restriction that is lacking in questions. [47] The second type of exclamative sentence may be divided into two subtypes: (57)–(59) vs. (60). The former consist of a noun phrase in the accusative, in most cases with an evaluative adjective. In the latter, as in (60), we find a noun phrase in the accusative and an infinitive (AcI). [48] In this subtype, the noun phrase often contains a deictic or quantifying Attribute, hardly ever [49] an evaluative adjective. The particle -ne does occur in the AcI, but hardly ever in the former subtype. Expressions such as edepol do not occur in the AcI, but are found in the former subtype. [50]

10.2.2 Illocutionary force

A serious problem for linguists who distinguish illocutionary forces is formed by the question of how many different illocutionary forces are to be distinguished and what criteria may be used to define each of them. I now address these two points ( crosssection 10.2.2.1). Then, for the first three sentence types distinguished on p. 191, I give a short survey of the illocutionary forces with which they occur ( crosssection 10.2.2.2).

10.2.2.1 Criteria for the distinction of sentence types and illocutionary forces Above I have already mentioned the occurrence of certain kinds of particles

-- 203 --

and words of other categories with certain sentence types or subtypes. In Latin linguistics such phenomena are occasionally pointed out, particularly in chapters on the various uses of the subjunctive and the differences between subjunctive and imperative. Also other criteria for distinguishing certain shades of meaning are mentioned. See example (61):

(61) quaeso hercle abire ut liceat :: abeas, si velis (`I am asking, by George, whether I may leave :: leave, if you want', Pl. Rud. 834)

The preceding modal auxiliary makes clear that abeas is not strictly directive. [51] See also example (31) above (p. 198), where debuisti makes clear how the ensuing rettulisses is to be understood.

Above I have limited myself to the question of which particles and the like can be combined with certain sentence types; we will see now, however, that the occurrence of such particles, parenthetical verbs, etc. is not only determined by the sentence type, but often also by the illocutionary force of sentences. For example, we have seen that quin occurs in imperative sentences with a directive illocutionary force (see example (42) on p. 199 above). We also find it in interrogative sentences with this illocutionary force, e.g. (62):

(62) quin tu taces? (`Will you be silent?', Pl. Men. 561)

I now discuss the various `tests' available for determining the illocutionary force of a sentence. These are the following:

(a) parenthetical or postposed verbs

(b) coordination and question/answer patterns

(c) modal particles

(d) restrictions other than modal particles

(e) data with regard to the situation (e.g. knowledge of the social class of the interlocutors)

(a) Parenthetical and postposed verbs

(63) per dexteram tuam te … oro obsecro, da mihi hanc veniam, ignosce, irata ne sies (`By your right hand, I beseech you, grant me this, forgive, do not be angry', Pl. Am. 923–4)

(64) tu quaeso cogita (`Please, think!', Cic. Att. 9.17.2)

For examples see K.–St. I. 199–202. [52] In recent publications on illocutionary force much attention is devoted to such parenthetical verbs, which, owing to their lexical meaning, make the intention of the speaker explicit, as a means to distinguish illocutionary forces (and often also sentence types). [53]

The verbs occurring parenthetically (always in the first person singular of the present tense) [54] can be subdivided into two groups:

(a) verbs meaning `to influence', `to appeal to': peto, precor, quaeso, obsecro, oro, rogo, obtestor, moneo; 54a

-- 204 --

(b) verba declarandi: dico, inquam, narro, fateor, confiteor, concedo (material: Plautus, Terence, Cicero, Petronius) and verba sentiendi (e.g. puto, credo, (ut) opinor (see above p. 192).

Most of these verbs do not occur very often, and not all four authors mentioned above have all of them. Of those that do occur, quaeso, obsecro, dico and inquam are highly frequent and idiomatic. [55] Consequently, they are found with all three sentence types and with all illocutionary forces, e.g. example (65) with quaeso, in a declarative sentence with an interrogative illocutionary force:

(65) quaeso … domina, certe embasicoetan iusseras dari (`Madam, I ask you, you surely had given orders that a lewd mug be presented to me', Petr. 24 (translation M. Heseltine, Loeb edition))

(b) Coordination; question/answer patterns

The illocutionary force of a sentence can sometimes be deduced from an immediately preceding or following sentence, see examples (66)–(67):

(66) non par videtur neque sit consentaneum (`It is not right and ought not to be done', Pl. Bac. 139)

(67) cynicum esse egentem oportet parasitum probe: ampullam … habeat (`A parasite ought definitely to be a needy cynic: he should have a bottle', Pl. Per. 123–5)

In (66) sit is a so-called potential subjunctive. In my terminology, there are two coordinated declarative sentences here. Habeat in (67) might be interpreted as jussive (and the sentence, thus, as directive), in view of the preceding oportet. [56] See also examples (31) and (61) on p. 198 and p. 203, respectively.

(c) Modal particles

Several times already we have encountered modal particles. K.–St. I. 199-202 mention as examples of modifying particles with imperatives sis, modo, dum, proin. Table 10.2 presents a survey of a number of particles and more or less idiomatic verb forms that occur in sentences with a directive illocutionary force. They do not, however, occur in each sentence type (see in particular sis and sodes). Table 10.2 Combinations of sentence types and particles and other expressions
ImperativeDeclarativeInterrogative
age+++
amabo+++
sodes+-+
sis+--
proinde+++
quin+++
Source: Bolkestein 1977c: 63.

-- 205 --

(d) The presence or obligatory absence of other kinds of constituent in certain sentence types has occasionally been mentioned above, e.g. restrictions on tense, addibility of Disjuncts (p. 197); modal verbs such as debere, licet are limited to interrogative and declarative sentences and embedded predications (which can then have a directive illocutionary force). [57]

(e) Data with regard to the situation

We have seen above that according to some scholars the choice between dixerit and dicat is determined by social relations (p. 194). Also other situational factors can influence the choice between alternative expressions. Thus, for instance, urgency can be a factor in choosing a short, matter-of-fact order (Stop thief!) or the desire to offer additional incentive (do your best). For Latin such factors have not been studied. [58] 58a

Although it is undoubtedly possible to develop more precise criteria to distinguish more subtly all kinds of nuances, one may wonder whether this will in fact lead to clearly distinguished illocutionary forces. It might be more useful to assume a gradation of illocutionary forces, ranging from, for instance, lexically expressed, explicit orders to cautious suggestions, e.g.:

(68) iubeo te abire (`I order you to go away': statement of an order)

(69) abi (`Go away')

(70) abeas si vis (`Go away if you want') [59]

10.2.2.2 Relation between sentence type and illocutionary force

In this section I restrict myself to the three illocutionary forces assertive, request for information and directive. Table 10.3 indicates which sentence types can occur with which illocutionary force. The imperative sentence type is the most limited one. It is not clear whether declarative sentences can be used with the illocutionary force `request for information'. See (65) on p. 204 above and (71):

(71) certe patrem tuum non occidisti (`You certainly did not kill your father, did you', Suet. Aug. 33.1)

Suetonius tells us that Augustus asked a question (ita fertur interrogasse (`Thus he is said to have asked')) and it is implied that the addressee answered (in casu no). We do not know whether (71) is a normal declarative sentence or was pronounced with a clearly interrogative intonation (as is the English translation). [60] Research should be done in order to determine how declarative sentences with a directive illocutionary force and imperative sentences differ semantically, and in what circumstances and text types which expressions are chosen.

-- 206 --

Table 10.3 Combination of sentence type and illocutionary force
Illocutionary force
Sentence typeAssertiveRequest for informationDirective
Declarative+?+
certe non … ? (71)ind. pres. pro imperativo (72)
ind. fut. pro imperativo (73)
Interrogative+++
so-called rhetorical question (74)(75)–(76)
Imperative--+

(71) certe patrem tuum non occidisti?

(72) itis, paratis arma quam primum viri (`Come on, men, prepare yourselves for battle as quickly as possible', Trag. inc. frag. 34)

(73) tu tamen …, ut adhuc fecisti, nos consiliis iuvabis (`You must help me with your advice, as you have done up to now', Cic. Att. 10.2.2)

(74) numquid Pomponius istis audiret leviora, pater si viveret? (`Would P. be listening to lighter things than these, if his father were alive?', Hor. S. 1.4.52–3)

(75) quin tu is accubitum? (`Won't you recline?', Pl. Ps. 891)

(76) non tu abis? (`Go away', Pl. Men. 516) [61]

10.3. The relation between the various uses of the moods

Above I have already pointed out that the current explanations for the various uses of the moods are based on the alleged semantic value of the individual moods itself and that they do not sufficiently take into account the various sentence types and the illocutionary forces mentioned above. A problem encountered by such an approach is how to distinguish the diverging semantic values of a certain mood in independent sentences and how to explain the relation between them. I first go into these two questions ( crosssection 10.3.1). A second issue is the relation between the use of the moods in independent sentences and in embedded predications. Something has already been said about this in crosssection 7.1.2. (pp. 100 f.). This is elaborated upon in crosssection 10.3.2.

-- 207 --

10.3.1 Distinction and relation between the uses of the moods

Difficulties in explaining the various uses of the moods especially occur in the case of the subjunctive, which, as we have seen, occurs in all three sentence types (declarative, interrogative, imperative). I will limit myself to the subjunctive.

In describing the use of the subjunctive, some linguists take as a starting-point the intepretations of the subjunctive in the contexts in which it occurs. More or less comparable interpretations are organized as much as possible and then labelled as a certain use. This approach resembles the treatment of the cases (see p. 63). A representative of this approach is K.–St. Contextual elements are frequently adduced as evidence, as we have seen above in crosssection 10.2.2. Besides constituents in the context, the translation to the native language of the linguist often plays an important role. Handford, for instance, supports his view that at one stage an element of the meaning of the subjunctive was that of `intention' with the statement that `there are other passages where the translation `I will' gives a perfectly good sense' (1946: 39). [62]

A second group of linguists explains the Latin uses as much as possible from the origins of the subjunctive. A representative of this approach is Sz. I first discuss one such diachronic explanation. Then I return to some synchronic views.

(i) The origin of the Latin subjunctive receives, and has received, much attention (see Sz. 329–30). [63] Scholars agree that the Latin subjunctive is used in situations in which, for example, Greek would use the optative and the subjunctive, while there are also many formal similarities. From this it is concluded that the Latin subjunctive is a merger of the Indo-European optative and subjunctive. [64] This leads Sz. to organize (with some reservations) his chapter on the subjunctive as follows:

(a) uses of the old optative: wish, concession, irrealis (330–5);

(b) uses of the old subjunctive: jussive, prohibitive, deliberative (335–8). [65]

The difference in negation (ne/non), as we have already seen, does not correlate with this classification.

(ii) The views described hereafter do not involve diachronic considerations. K.–St. distinguish there uses of the subjunctive, viz. the `potential' subjunctive (I.176–80), the `volitive' subjunctive (I.180–95) and the `conditional' subjunctive (I. 195, limited to imperfect and pluperfect subjunctive forms). They argue that all these uses can be explained from a common denominator, viz. that all `present something as thought, as imagined' [66] (see Sz. 326: `Modus der Vorstellung'). Opposed to this are the indicative and the imperative as `Modus der Wirklichkeit' (`mood of reality') and `Modus des Befehls' (`mood of orders'), respectively. Others distinguish two uses, viz. the volitive subjunctive (negation ne) and the non-volitive subjunctive (negation non), for

-- 208 --

which no common denominator can be found. [67] A third approach explicitly considers the various uses as contextually determined within a kind of continuum of meanings that cannot very well be captured by a common denominator. [68] The unity of the subjunctive consists in its being opposed to the indicative and the imperative. 68a

To some extent, my own view has already been expressed on p. 194. I observe that in declarative sentences there is a semantic opposition e.g. (non) facio/faciam/facerem. I have described this difference with the terms `factive', `possible', `purely hypothetical', respectively. We find the same difference in interrogative sentences. In imperative sentences there is no semantic opposition (ne) facit/faciat, but there is an opposition (ne) faciat/faceret. In this case there is a contrast between possible and hypothetical. This has been represented in table 10.4 (see also table 10.1 on p. 190.) [69] Table 10.4
Mood
Sentence typeIndicativeSubjunctive I (present/perfect)Subjunctive II (imperfect/pluperfect)
Declarative (non)+++
Interrogative (non)+++
Imperative (ne)-+(+)
An indication for the correctness of table 10.4 may be found in a number of properties of interrogative sentences in indirect speech. The subjunctive I and subjunctive II distinguished in table 10.4 are differentiated in indirect speech: indicative → infinitive, subjunctive I → subjunctive, subjunctive II → infinitive construction with -urum fuisse or infinitive + potuisse, see (77) – (79): [70]

(77) quonam haec omnia nisi ad suam perniciem pertinere? (`To what did all this lead, apart from his own downfall?', Caes. Civ. 1.9.4) [cf. quonam haec pertinent]

(78) quis hoc sibi persuaderet sine certa spe Ambiorigem ad eiusmodi consilium descendisse (`Who would make himself believe that A. had developed this plan without certain expectations?', Caes. Gal. 5.29.5) [cf. quis persuadeat]

(79) quid illum facturum fuisse si … adversa pugna evenisset (`What he would have done if the fight had finished badly', Liv. 8.31.5) [cf. quid fecisset]

-- 209 --

10.3.2 The moods in independent sentences and in subordinate clauses

In chapter 7 we have discussed restrictions on controllability and tense of predications with an imperative modality, both embedded and independent predications. This was also pointed out in this chapter (p. 197). In imperative embedded predications with a finite verb from the mood is, as we have seen, the subjunctive (and the negation is ne). An example of such an embedded predication is (80):

(80) hortatus sum ut ea quae sciret sine timore indicaret (`I have encouraged him to set forth what he knew without fear', Cic. Catil. 3.8)

I start with argument clauses, where the mood is fixed and can usually be explained. Then I discuss satellite clauses where the mood is generally also fixed. Finally clauses are treated where a `choice' is involved.

(i) With all kinds of verbs meaning `to take care of', `to strive after', `to advise', `to order', `to decide', `to allow', `to wish', the embedded predication required by the verb (an argument in the sentence, therefore) cannot possibly be factive. It is, therefore, in a way `consistent' [71] that-besides the similarity in terms of restrictions mentioned above-also in terms of mood there is a correlation between subordinate clauses and independent imperative sentences: there, too, the `non-factive' mood is common. With verbs such as gaudere and dolere, on the other hand, where the embedded predications does refer to something that is actually the case, the indicative is what one expects in the quod-clause. See example (81):

(81) sane gaudeo quod te interpellavi (`I am certainly happy to have interrupted you', Cic. Leg. 3.1)

Note that in embedded predications with verbs like gaudere the negation is, again, non and that restrictions of the kind mentioned above do not exist. Note also the difference in subordinator (ut/quod). But the `consistency' mentioned just now is lacking with other classes of verbs, e.g. verbs meaning `to realize' and `to happen'. With the former class of verbs, we find in the embedded predication both ne and non as negation (See K.-St. II.212–3; 234–5). With the verbs meaning `to happen' the subjunctive is normal (with non as negation), although the embedded predication frequently contains a state of affairs that has really taken place. An example is (82):

(82) mihi … valde optanti … cecidit ut in istum sermonem … delaberemini (`To my pleasure it so happened that you ended up in this discussion', Cic. de Orat. 1.96) [72]

Note that in none of the examples given above can the subjunctive be replaced by the indicative or vice versa.

In indirect questions Classical Latin uses the subjunctive. In Plautus, on the other hand, the same governing verbs seem to allow also the indicative. Closer

-- 210 --

examination shows, however, that in the instances with the indicative the governing verbs are rather used parenthetically, so that the questions are direct questions, where the indicative is normal. There is, however, no semantic motivation for the occurrence of the subjunctive in real indirect questions. [73]

(ii) Also in most satellite clauses the mood is fixed. It is closely connected with the subordinator that is chosen (or, in the case of very frequent subordinators, with certain meanings of the subordinator). In the majority of satellite clauses the negation is non (in Purpose clauses it is ne): [74]

(83) quam (provinciam) iste … ita vexavit ac perdidit ut ea restitui in antiquum statum nullo modo possit (`He has destroyed and ruined this province in such a way that it cannot in any way be restored to its original state', Cic. Ver. 1.12)

As in (83), the content of the embedded predication is in many cases `fa–ctive', so that–on the basis of the semantic value of the subjunctive given above–the subjunctive cannot be explained. Nor can it be explained in instances such as (84)–(85):

(84) nec eorum quisquam adhuc inventus est cui quod haberet esset satis (`And among them nobody has as yet been found for whom what he had was enough', Cic. Parad. 52)

(85) est oratoris, quaecumque res infinite posita sit, de ea posse dicere (`An orator must be capable of speaking about any general subject that is put before him', Cic. de Orat. 2.66)

In (84) we find a so-called second-degree subordinate clause, with attraction of the mood as a result of the mood of the governing clause (esset). (85) is an example of the rule that in subordinate clauses with infinitive and AcI constructions the subjunctive is often used. In the same connection, the subjunctive is used in subordinate clauses in indirect speech (`oratio obliqua') (negation non). All these instances ((82)–(85)), where the subjunctive cannot be accounted for on the basis of its alleged semantic value in independent sentences, must be explained in a different way: at least for some time the subjunctive developed into a formal means to mark subordinate clauses as such; in other words, it was a morpho-syntactic rather than a morpho-semantic means. [75]

Above, on p. 194, I have pointed out that in independent sentences the difference between present subjunctive and imperfect subjunctive is in a sense `atemporal'. In embedded clauses, however, there is generally a temporal difference, visible in, among other things, the so-called sequence of tenses. [76]

(iii) Up to here we have discussed subordinate clauses where no opposition between subjunctive and indicative is possible. In a certain number of

-- 211 --

subordinate clauses (satellite clauses and relative clauses), however, there is a semantic difference between clauses in the indicative and in the subjunctive. I first deal with satellite clauses, then with relative clauses.

In a limited number of subordinate clauses, viz. those introduced by the subordinators quod, priusquam, dum we find both the indicative (the (a)-examples) and the subjunctive (the (b)-examples). See (86)–(88):

(86a) urbs quae, quod in ea parte Fortunae fanum … fuit, Tycha nominata est (`A city that is called tycha, because nearby there was once a temple of Fortuna', Cic. Ver. 4. 119)

(86b) Aristides … nonne ob eam causam expulsus est patria, quod praeter modum iustus esset (`Hasn't A. been expelled from his country, because–according to his fellow-countrymen–he was excessively just?' Cic. Tusc. 5.105)

(87a) quod ego … priusquam loqui coepisti … sensi (`As I noticed before you began to speak', Cic. Vat. 4)

(87b) saepe magna indoles virtutis, priusquam rei publicae prodesse potuisset, extincta est (`Often great innate talents are destroyed before the state could benefit by them', Cic. Phil. 5.47)

(88a) perpaucos dies, dum pecunia accipitur … commorabor (`I shall stay for a few days, until the money has arrived', Cic. Fam. 3.5.4)

(88b) subsedi in ipsa via dum haec … summatim tibi perscriberem (`I am sitting in the street while briefly writing this to you', Cic. Att. 5.16.1)

For examples see K.–St. (quod: II.383 ff.; priusquam: II.366 ff.; dum: II. 380 ff.). The subjunctive in (86b) and (87b) can be explained because the predications are clearly non-factive, while in the corresponding examples with the indicative, on the other hand, the predication is factive. But the subjunctive in (88b) is more difficult. The dum-clause seems interchangeable with a Purpose ut-clause and indicates the intention of the Subject of subsedi. Presumably, Cicero is writing at that very moment.

The relative clauses with the subjunctive can be subdivided into various types. [77] First, the so-called final relative clauses. Examples are (89)–(91):

(89) delegisti quos Romae relinqueres, quos tecum educeres (`You chose people that were to be left at Rome and people to take with you', Cic. Catil. 1.9)

(90) praeterea se missum a M. Crasso qui Catilinae nuntiaret … (`That, moreover, he had been sent by M.C. in order to report to C. … ', Sal. Cat. 48.4)

(91) oriens incendium qui restinguerent summos viros misimus (`We have

-- 212 --

sent very important men in order to extinguish the rising fire', Cic. Phil. 13.48)

In these instances, the events described in the relative clauses are posterior to those described in the main predication, and are to be brought about by certain persons. The states of affairs of the main predication and the relative clause are Actions (controllable and dynamic). In those cases where the antecedent of the relative pronoun is present in the main predication the relative clause could be replaced by an ut-clause. [78]

The following instances would (presumably) be regarded as `generic' or `consecutive' by the grammars: [79]

(92) ego enim suscipiam et, ut spero, reperiam qui id quod salutis omnium causa statueritis, non pute <n> t esse suae dignitatis recusare (`For I shall undertake this and hope to find many people who regard it as undignified to refuse what you have decided in the interest of all', Cic. Catil. 4.8)

(93) multa … e corpore existunt, quae acuant mentem (`In the body many things occur that sharpen the mind', Cic. Tusc. 1.80)

(94) cum haec essent ita constituta, Verres tot annis atque adeo saeculis tot inventus est, qui haec non commutaret sed everteret (`When things had been arranged in this way, years, even centuries later (a) Verres appeared, not simply to change the arrangement, but to overturn it', Cic. Ver. 3.21)

In these examples, the state of affairs of the main predication is non-controllable. The relative clauses contain states of affairs that will possibly or probably obtain. A common element in both groups of examples (the so-called final and consecutive relative clauses) is the non-factivity of the predications of the relative clauses.

This is not the case in the following group of examples:

(95) Paetus … omnes libros quos frater suus reliquisset mihi donavit (`P. has given to me all the books that his brother-as he said-had left him', Cic. Att. 2.1.12)

(96) illi autem, qui omnia de republica praeclara atque egregia sentirent, sine recusatione … negotium susceperunt (`But they, endowed with excellent and extraordinary feelings with regard to the public interest, accepted the task without hesitation', Cic. Catil. 3.5)

In (95) Cicero does not express himself as to the truth of the fact that `all books were left by P.'s brother', but presents this as P.'s view. In (96) Cicero does not doubt the good faith of illi, but indicates by means of the subjunctive that their behaviour was a logical outcome of their feelings. Sentiebant, indicating that they had those feelings at that moment, would be either trivial or even

-- 213 --

insulting, since it would suggest that they no longer had those feelings. So, while the first group (the final and consecutive relative clauses) involves non-factivity, here a factive expression is, as it were, avoided.

The following instances are related to these:

(97) vehementer Sullam probo, qui tribunis plebis sua lege iniuriae faciendae potestatem ademerit (`I appreciate Sulla very much, since by his law he has deprived the tribunes of the possibility of committing acts of injustice', Cic. Leg. 3.22)

(98) egomet qui sero ac leviter Graecas litteras attigissem tamen cum … venissem Athenas, complures tum ibi dies sum … commoratus (`Even though I had come to know Greek literature relatively late in life and in a superficial manner, I nevertheless stayed in Athens for a few days after my arrival', Cic. de Orat. 1. 82)

(97) and (98) are examples of a so-called causal relative clause and a so-called concessive relative clause, respectively. In the former case, the context contains indications as to the interpretation, e.g. the evaluative verb probo; [80] in (98) the concessive interpretation is supported by tamen. In both instances the indicative would also be possible. [81] Yet, the subjunctive gradually becomes more frequent. Note that the clauses are non-restrictive (the antecedents are definite). This was not (always) the case in the previous instances. Note that (97) and (98) involve events that have in fact taken place. Here, too, a lot of research remains to be done.

Bibliographical information

On sentence types in Latin few general studies have appeared, see Rubio (1968) and Scherer (1975); in connection with illocutionary force also Bolkestein (1980a: chapter 5). General studies on illocutionary force are, apart from Austin (1962), who introduced the notion, Franck (1980: chapter 3), Lyons (1977: chapter 16) and Searle (1969; 1976). More research has been done on individual sentence types. For exclamative (and interrogative) sentences (in Plautus) Vairel-Carron (1975) can be recommended. Imperative sentences, though not described in terms of sentence types, are treated elaborately in Löfstedt (1966) and Vairel-Carron (1975). Very much has been written on the moods. Exhaustive surveys of the literature are to be found in Calboli (1966; 1968; 1983). A survey of the uses of the moods is Handford (1946). Many problematic issues are discussed by Thomas (1938). An attempt to formalize the different uses of the subjunctive is Lehmann (1973). Touratier (1977; 1983b) offers recent surveys of the various views on the moods. The approach I have chosen in this chapter is closest to that found in Rubio (1968).

-- 214 --

Previous Section

Next Section


Pinkster, Harm (1942-) [1990], Latin Syntax and Semantics [info], xii, 320 p.: ill.; 24 cm. [word count] [Pinkster].
Powered by PhiloLogic