Navigation – Plan du site
Le verbe

Auxiliaries in double-negation settings (non potest non + inf): Semantic gliding and syntactic description

Les auxiliaires dans le cadre de la double negation (non potest non + inf) : glissements sémantiques et description syntaxique
Adriana M. Manfredini
p. 117-125

Résumés

L’expression non potest non + inf véhicule un sens de nécessité qui reflète la relation d’implication entre possibilité et nécessité. Les données examinées montrent que cette implication est plus strictement orientée vers le sous-domaine déontique, en prenant en considération des traits sémantiques de la prédication de l’infinitif. Et en fin de compte, l’analyse prouve que cette tournure peut être interprétée comme l’expression grammaticale de la possibilité exclusive.

Haut de page

Texte intégral

I would like to thank the anonymous reviewers for their corrections and suggestions, which I have taken into account; any remaining faults are my own responsibility.

1. The logical equivalence

1Examples like those in (1) convey a meaning of necessity, although it is lexically expressed by means of possum.

(1)

a.

Itaque Gallus Aelius ait: “Inter legem et rogationem hoc interest. Rogatio est genus legis; quae lex, non continuo ea rogatio est. <Rogatio> non1 potest non2 esse lex, si modo iustis comitiis rogata est.” (Fest., p. 266)
‘Consequently Gallus Aelius says: “This is the difference between a law and a bill. A bill is a kind of law; if something is a law, it is not necessarily a bill. It is not possible for a bill not to be a law, provided that it has been approved in a legitimate assembly.”’

b.

Nam ceteri dolores mitigantur vetustate, hic non1 potest non2 et sensu praesentis miseriae et recordatione praeteritae vitae cottidie augeri. (Cic. Att. 3, 15, 2)
‘For other griefs are lessened with age, this one cannot but increase day by day with my feeling of present misery and my memory of life gone by.’

c.

Vir bonus non1 potest non2 facere quod facit; non enim erit bonus, nisi fecerit… (Sen. Benef. 6, 21, 2)
‘It is not possible for a good man not to do what he does; for he will not be good, unless he has done so…’

2This is a special case of double negation where two negative operators have scope on different lexical entities, and even so, the outcome is an affirmation. In all three cases, the logical equivalence is the one expressed in (2) through a pair of duals, in which “what is necessarily the case is impossible not to be the case, and vice versa. So too in deontic (obligation and permission-based) logics, an act is permissible by definition if there is no obligation not to perform it”, says Horn (1978, p. 163), and I add, vice versa as well:

  • 1 nec = necessity; poss = possibility; p = proposition.

(2)

nec p ≅ ~ poss ~ p (cf. Bertocchi and Orlandini, 2002, p. 10).1

3There is an implication relationship between necessity and possibility which is responsible for the fact that “double negatives of the logical form ‘not V not’ will thus result in stronger assertions than V’, if V is the weaker (entailed) member of the dual opposition: but weaker assertions than V’, if V is the stronger (entailing) member” (Horn, 1978, p. 164). Accordingly, all the examples in (1) are cases of strong assertion, since the weaker member of the opposition (posse) is being used in the double-negation setting to signify debere.

4Linguistically speaking, necessity and possibility can encompass diverse modal values: epistemic, deontic, and dynamic. Following Palmer (2001), epistemic modality conveys the speaker’s judgement over the whole proposition, whereas deontic and dynamic modalities refer to the event, and are essentially prospective, in the sense that the event they relate to is not yet actualized (Palmer, 2001, p. 70). In deontic modality, an obligation emanates from an external source, whereas in dynamic modality an internal condition of the subject is relevant for the completion of the event (Magni, 2010, p. 198). Schematically:

  • 2 For a formal difference between future will and volitive will, cf. Palmer, 2001, p. 78. I follow wh (...)

(3)

You must go.

deontic: ‘You have the obligation to go.’

You can come in.

deontic: ‘You have permission to come in.’

(4)

He must be in his office.

epistemic: ‘I conclude that he is in his office.’

He may be in his office.

epistemic: ‘It is possible that he is in his office.’

(5)

He will help you.

dynamic: ‘He has the willingness to help you.’2

He can run a mile.

dynamic: internal condition: ability of the subject – ‘He has the ability to run a mile.’

He can escape.

dynamic: external circumstances: the door is not locked – ‘He has the possibility to escape.’

  • 3 For now, it is not my intention to discuss the implications of this representation for auxiliary ph (...)

5These values, though so simply exemplified in (3)-(5), are not so easily distinguishable in the cases included in (1) because they do not come out straightforwardly, being submitted to the domain of negation through the logical dual opposition relationship shown in (2). What can be inferred from Horn’s words quoted above is that the logical formula is semantically neutral as regards its application to epistemic or deontic meanings. At first sight, the concrete linguistic expression seems to mirror the logical formula: the examples in (1) make clear that each of the negative operators has differentiated scopes: non1 falls on the modal auxiliary, whereas non2 affects the non-finite predication. In the literature, non1 is also referred to as external negation and non2 as internal negation (for instance, Bertocchi and Orlandini, 2002), or in more transparent terminology, auxiliary negation and main verb negation, respectively (cf. Quirk et al., 1985, § 10.67). This can be diagrammed as follows:3

(6)

[ [non1 [posse]] [ non2 [infinitive main predication] ] ]

  • 4 In Cormack and Smith, 2002, each negative polarity item is said to be located on a different level: (...)

6Nevertheless, even if it is conceived as a sort of hierarchical distribution as proposed in (6), where non1 + possum has domain over the infinitive main predication under the scope of non2, this syntactic arrangement cannot be solely responsible for a more or less definite value of ‘necessity’.4 It is the semantico-syntactic structure and the contextual and discourse information that cooperate to determine the corresponding modal meaning, although no claim can be made that this is unequivocal or unambiguous.

7This paper will mainly try to determine which of the diverse senses of necessity this linguistic formula is compatible with, and what features can be more or less conditioning in order to establish the different values, once the semantic gliding from possibility has been accomplished through the intercession of negation.

8While it is quite common to oppose both deontic and dynamic meanings as agent-oriented against epistemic meanings as non agent-oriented, this distinction does not prove to be completely determinant: if that were the case, examples (1a) and (1b) should be considered clearly epistemic, but as it will be shown later on, there are reasons to read them as deontic. As regards (1c), a quick look might raise the question whether its deontic (agent-oriented) meaning does not conceal a dynamic interpretation: it is the subject’s internal condition of goodness that necessarily becomes willingness.

2. Epistemic vs. deontic

9In a paper on Dutch modals, Barbiers (2002, p. 58ff.), discusses the differences between root interpretation (what could be here quickly identified with dynamic and deontic meanings) and epistemic interpretation, and offers the following cases:

(7)

De aarde moet een planet zijn.
‘It must be true that the earth is a planet.’

(8)

God said: ‘The earth must be a planet.’
‘God required that the earth be a planet (and not a star).’

10Barbiers claims that when the infinitive main predication denotes a fixed property of the subject, the reading is epistemic, but if it denotes a variable one, the interpretation is deontic. In (7), given the facts of the real world as it is, Barbiers accepts only the epistemic reading; in (8), in a possible world, the earth is not yet a planet, and so a deontic reading (root interpretation) is derived. In this last case, Barbiers argues that there exists what he calls a polarity transition: “(i) the complement of the modal must denote a value on a bounded scale, and (ii) this value is not the actual value at the moment denoted by the modal” (Barbiers, 2002, p. 59). While analyzing the semantic properties of infinitive predications, Bolkestein (1980, p. 63-4), states that epistemic modal meanings apply to “states of affairs conceived of as actually existing at some point in time previous, co-temporaneous or future to the time of speech”; in contrast, deontic modal meanings are related to states of affairs that do not convey a deictic temporal reference, and are for this reason “potential” (her quotation marks). Having all these considerations in mind, I propose a more thorough examination of example (1a), which I repeat and renumber here:

(9)

Itaque Gallus Aelius ait: “Inter legem et rogationem hoc interest. Rogatio est genus legis; quae lex, non continuo ea rogatio est. <Rogatio> non1 potest non2 esse lex, si modo iustis comitiis rogata est.” (Fest., p. 266)
‘Consequently Gallus Aelius says: “This is the difference between a law and a bill. A bill is a kind of law; if something is a law, it is not necessarily
a bill. It is not possible for a bill not to be a law, provided that it has been approved in a legitimate assembly.”’

11It is true that a rogatio is a law, but the converse does not hold. In order to be a law, the external condition is that it emerges from a legitimate assembly, which represents the normal, regular, juridical procedure, as is stated in the protasis. The perfect tense in the conditional clause simply shows that this must be fulfilled before a rogatio is considered as such, but it is in no sense an actual time reference to a concrete rogatio. This recommends a deontic reading of the passage.

12In a similar way, then, the same deontic intention can be found in Cicero’s words in (1b), since his grief is contrary to the nature of other kinds of pain, because it grows with daily feeling and recollection. This is an inevitable condition of Cicero’s sorrow and it is this very feature which makes it a prospective, non-fixed property.

13In both of the previous examples, there is an external legal or natural frame to which the events of the propositions are referred. Among the features that cooperate to give Spanish deber an epistemic meaning, which she terms “contextual”, as distinct from an “invariable meaning” – a “core” sense: ‘requires X’ –, Silva-Corvalán (1995, p. 88-9), proposes that one of the requirements is the speaker’s neutral attitude towards the proposition. But this does not seem to be the case in (9), since speaker’s attitude is neutral, but the passage results in a deontic reading nonetheless. This is in contrast to Cicero’s attitude in (1b) and also in the next example (10), where Cicero seems more invested:

(10)

Tale enim tuum iudicium de homine eo quem ego unice diligo non potest mihi non summe esse iucundum; quod cum ita sit, esse gratum necesse est. (Cic. Fam. 13, 18, 2)
‘For such an opinion on your part of a man who stands alone in my affection cannot but give me the highest degree of pleasure. And that being so, I cannot help but be grateful.’

  • 5 The anonymous reviewers remark both examples (9) and (10) have epistemic reading: my own explanatio (...)

14Cicero’s affection for his friend Atticus is a fact well-known to the addressee of the letter (Servius Sulpicius), and Cicero’s previous praise of Servius’ politeness and favors towards Atticus constitute the discourse background against which an obliged feeling of gratitude takes the stage. This statement about his personal delight is clearly prospective, not present-bound, since it consequently entails gratitude as a naturally derived feeling, and so he writes necesse est. What provokes the deontic nuance in the passage is the set of socially regulated principles comprised: Servius’ opinion is a pleasant one, because it is Cicero’s amicitia with Atticus that is being well judged, after all. Gratitude is the normal consequence of being happily favoured.5

15This next example also suggests a deontic reading. At first sight, since these are Cicero’s compliments to M. Marcellus, as the previous context let us understand, the evidence brought up in the causal quod clause justifies the conclusion that Marcellus should be praised by all:

(11)

(Itaque neque tu multum interfuisti rebus gerendis et ego id semper egi ne interessem. Non enim iis rebus pugnabamus quibus valere poteramus, consilio, auctoritate, causa, quae erant in nobis superiora, sed lacertis et viribus, quibus pares non eramus…) In quo tuum consilium nemo potest non maxime laudare, quod cum spe vincendi simul abiecisti certandi etiam cupiditatem ostendistique sapientem et bonum civem initia belli civilis invitum suscipere, extrema libenter non persequi. (Cic. Fam. 4, 7, 2)
‘(You, therefore, took no great part in the campaign, and I always made a point
of taking no part at all. For we were not fighting with the weapons which might have given us strength, such as judgement, the weight of personality, or the soundness of our cause, in all of which we were superior, but with the brute force of our muscles…) And here it is impossible for any man not to commend most cordially your decision, as soon as you saw there was no hope of victory, to cast out of your heart every desire to continue the struggle, proving thereby that a wise man and honest citizen, while he hesitates to be responsible for the inception of a civil war, has no hesitation in refusing to carry it through to the bitter end.’

16The first implication is that everybody must praise Marcellus (note Cicero’s approval through the mention of his own rejection to war in the previous context). Given that war is not being positively evaluated against some other highly renowned virtues (consilium, auctoritas, causa), these external conditions suggest that an ethical norm is also at stake, so that a deontic interpretation is certainly inevitable. In this last case, the praise is something yet to come, not a certain and actualized state of affairs.

17So far, it is not agentness that is fully compromised in deontic readings, as could be the case in (11), but internal or external conditions present in the discourse context: for a deontic interpretation, there is an external circumstance which justifies the obligation that must be accomplished.

  • 6 Perception verbs also have experiencer subjects and, according to Magni, 2010, p. 218, they can tri (...)

18In the next three cases, subjects are experiencers of processes, states of affairs that should in principle trigger epistemic readings.6 But nothing of the kind occurs:

(12)

Non possum eius (sc. Pompeii) casum non dolere; hominem enim integrum et castum et gravem cognovi. (Cic. Att. 11, 6, 5)
‘I cannot help feeling sorry for his fate, for I knew him to be a man of honour
and high moral principle.’

(13)

Iuvenem nostrum non possum non amare, sed ab eo nos non amari plane intellego. (Cic. Att. 10, 10, 6)
‘My nephew I cannot but love, though I see clearly that he has no affection
for me.’

(14)

Sed tota res Brundisina quo modo habeat se diutius nescire non possum. (Cic. Att. 9, 3, 2)
‘But I cannot much longer ignore how the whole affair at Brundisium is going
.’

  • 7 See Bolkestein, 1980, p. 65, where she speaks of ‘hidden directive’, defined as a directive that “d (...)

19In (12), in fact Cicero feels sympathy for Pompeius’ death, but there is no real temporal reference to actual present: pity is an acquired feeling based on the integrity assigned to the deceased, and it is a sort of self-recommendation or indirect order.7 In (13), Cicero’s affection for his nephew has a peculiar proof through the negative evidence of the boy’s indifference to him. If this situation described by Cicero is taken as non-actual present, and is in itself understood as a natural condition between members of the same family, then a sense of deontic necessity cannot be denied. Of course, that the boy does not behave according to the same social standards is a proof that Cicero actually does.

  • 8 The only counterexample to this observation in my present corpus is Sen. Benef. 3, 19, 2.

20As regards (14), Cicero knows that Caesar is already in Brundisium, but does not know for sure whether he will go after Pompeius or not. In this context his ignorance is real, and it corresponds with Cicero’s present state of affairs, but also this personal circumstance is extending in time (diutius nescire). So, this non-perfective state conspires against an epistemic reading (there are external conditions: information he might be expecting from authorized sources is on its way). Besides, some other interpretation cannot be ruled out here: that Cicero does not allow himself to stay ignorant for such a long time. In other words, it looks like non possum = ‘no permission (prohibition)’, and nescire is a “positive” lexeme, in the sense that negation is a meaning already embedded in the word stem as a semantic feature, and is not a polarity operator. If this is true, word order could also provide support for this conclusion: a quick look at the data turned up so far shows that non posse non + inf. does not alter this sequence, whereas in cases when the modal auxiliary is the only one being negated, the infinitive may precede non posse (just like nescire non possum or, for instance: ubi honos publice non est, ibi gloriae cupiditas esse non potest, Cic. Agr. 2, 91).8

21Since this is not an exhaustive survey of data, the partial conclusion is that the formula non posse non + inf seems to convey deontic meanings, irrespective of semantic properties of subjects, and more in tune with the non-fixedness of the state of affairs or properties assigned to them; these factors together with contextual information have served to tip the scales towards an almost all-pervading sense of obligation.

22Even though possibility and necessity maintain logical connection, it is no less true that it can be seen, even from the translations attached to the texts that ‘possibility’ does not entirely disappear from sight. Again, the logical implication accounted for at the beginning of this paper could be invoked to provide an explanation for it, even for the fact that, after all, whenever we say non posse non + inf we also say ‘only x is possible’. This exclusive meaning also contains a sense of inevitability that pushes it towards necessity, in the sense that it reduces all outcomes to only one: no contingent possibility is contemplated.

23Bertocchi and Orlandini (2002, p. 21-2) propose that fieri non potest quin is the Latin expression for epistemic necessity. According to my data, non potest non + inf could then be considered more specific of deontic necessity.

3. Dynamic reading?

24Finally, I offer one last text – as a matter of fact, a larger passage that encompasses example (1c) at the beginning. A brief glance at (15) is sufficient to wonder whether the same expression is not capable of conveying a hybrid dynamic meaning of will and ability, not easily detachable from the deontic value of the formula:

  • 9 I reproduce the English translation by John W. Basore, The Loeb Classical Library, London, 1928-193 (...)

(15)

Sciam,’ inquit, ‘solem ac lunam nobis velle prodesse, si nolle potuerint; illis autem non licet non moveri. Ad summam consistant et opus suum intermittant.’ Hoc vide quot modis refellatur. Non ideo minus vult, qui non potest nolle; immo maximum argumentum est firmae voluntatis ne mutari quidem posse. Vir bonus non potest non facere quod facit; non enim erit bonus, nisi fecerit; ergo nec bonus vir beneficium dat, quia facit quod debet, non potest autem non facere quod debet. Praeterea multum interest, utrum dicas: ‘Non potest hoc non facere’ quia cogitur, an: ‘Non potest nolle.’ (Sen. Benef. 6, 21, 1-4)
‘“I should be sure,” you say, “that the sun and the moon really wish to do us a service if it was possible for them to be unwilling; but they cannot help being in motion. In short, let them halt and discontinue their work.” But see in how many ways this argument may be refuted. It is not true that a man who is unable to refuse
is for that reason the less willing to do; nay, the greatest proof of a fixed desire is the impossibility of its being altered. A good man is unable to fail to do what he does; for unless he did it, he would not be a good man. And, therefore, a good man gives a benefit, not because he does what he ought to do, but because it is not possible for him not to do what he ought to do. Besides, it makes a great difference whether you say: “It is not possible for him not to do this,” because he is forced to do it, or: “It is not possible for him to be unwilling.”’9

  • 10 For licet as a modal that conveys permission, see, for instance, Orlandini, 1998, p. 1023, n. 11, a (...)
  • 11 Note the change in word order, compared to the first instance at the beginning of this passage.

25The same doubt might arise with nolle potuerint as with nescire possum in (14) before; but, given the contrast with velle in the same line, it could be accepted that non posse nolle is actually being used by Seneca as a double negation; dynamic volition is the meaning implied, since we cannot deny the [+control] associated with nolle/velle. But the volition is regulated by an obligation, non licet non moveri, if we accept that there is the ‘prohibition of not doing x’.10 In this respect, the second instance of non potest nolle is ambiguous: on the one hand, it is deontic, because there is a determination not to change, in accordance with external norms; however, this very same determination is a sign of internal capacity on the side of the subject.11 So, it is an internal condition of the good man to do what he does, because he wishes to do so: this controllable situation is slightly different from the simple attachment to rules, or rather, it is the passage from the external adjustment to a norm to an internal norm, and that is the will. In the intent affirmation non potest autem non facere quod debet there is some inevitable capacity to do good implied as well.

26This last example is a valid proof of the fuzziness of what is entailed by the logical opposition diagrammed in (2). If a hint of dynamic sense can be perceived in most of the double negations in bold, it is triggered by the lexical items such as possum and volo/nolo, rather than by the strict logical equivalence or the dual opposites. The wording conveys necessity, but focuses on possibility.

4. Conclusion

27The entailment of possibility and necessity is more strictly directed towards the deontic subdomain through the compositional formula non potest non + inf, as a copy of the logical equation. But it is also clear that it seems to develop a more linguistic function, as a sort of grammatical structure that conveys an exclusive sense of possibility: ultimately, the utterance expresses the only possibility that is left for some subject or entity (whatever semantic subdomain of modal value is chosen). From this point of view, and given the alternation that can be seen in (15), non licet non moveri, the hierarchical structure drawn at the beginning of this paper in (6) would need one more step into grammatical abstraction, and be reformulated as is shown in (16), to become a grammatical structure of exclusive possibility.

(16)

[ [non1 [ auxposs] ] [ non2 +infinitive main predication ] ] ]

Haut de page

Bibliographie

Barbiers, S., 2002, Modality and polarity, in S. Barbiers et al. (eds.), Modality and its Interaction with the Verbal System, Amsterdam/Philadelphia, p. 51-73.

Bertocchi, A. and Orlandini, A., 2002, Impossible n’est pas latin. Sur les concepts modaux de “possible” et d’“impossible”, et leur réalisation en latin, in M. Fruyt and C. Moussy (eds.), Les modalités en latin, Paris, p. 9-23.

Bolkestein, M., 1980, Problems in the Description of Modal Verbs. An Investigation of Latin, Assen.

Cormack, A. and Smith, N., 2002, Modals and negation in English, in S. Barbiers et al. (eds.), Modality and its Interaction with the Verbal System, Amsterdam/Philadelphia, p. 133-163.

Di Tullio, A., 2014, Manual de gramática del español, Buenos Aires.

Giammatteo, M. and Marcovecchio, A. M., 2010, Perífrasis verbales: una mirada desde los universales lingüísticos, Sintagma 21, p. 21-38.

Horn, L., 1978, Some aspects of negation, in J. Greenberg et al. (eds.), Universals of Human Language, vol. 4: Syntax, Stanford, California, p. 127-210.

Magni, E., 2010, Mood and modality, in P. Baldi and P. Cuzzolin (eds.), New Perspectives in Historical Latin Syntax, vol. 2: Constituent Syntax: Adverbial Phrases, Adverbs, Mood, Tense, Berlin/New York, p. 193-275.

Manfredini, A., in press, Auxiliaries within comparative clauses: Some remarks concerning their syntax and grammatical description, in P. Poccetti (ed.), Latinitatis rationes. Descriptive and Historical Accounts for the Latin Language, Berlin / New York.

Orlandini, A., 1998, La polysémie du prédicat “pouvoir” et sa désambiguïsation en latin, in B. García Hernández (ed.), Estudios de lingüística latina. Actas del IX Coloquio Internacional de Lingüística Latina, vol. 2, Madrid, p. 1017-1031.

Palmer, F., 2001, Mood and Modality, Cambridge (2nd edition).

Quirk, R. et al., 1985, A Grammar of Contemporary English, London.

Silva Corvalán, C., 1995, Contextual conditions for the interpretation of poder and deber in Spanish, in J. Bybee and S. Fleischman (eds.), Modality in Grammar and Discourse, Amsterdam/Philadelphia, p. 453-471.

Haut de page

Notes

1 nec = necessity; poss = possibility; p = proposition.

2 For a formal difference between future will and volitive will, cf. Palmer, 2001, p. 78. I follow what I understand in his presentation regarding ‘willingness’ as the dynamic counterpart of ‘necessity’ in order to preserve his bipartite system idealization of the English modals.

3 For now, it is not my intention to discuss the implications of this representation for auxiliary phrases, since it is clear that negation is not interacting with a periphrasis: for Spanish epistemic poder, the admission of internal negation is taken by Di Tullio, 2014, p. 246, as a proof of the non-existence of auxiliary phrase. Though supporting the idea of a periphrastic unit, Giammatteo and Marcovecchio, 2010, argue that double negation is proof of a low level of auxiliary phrase grammaticalization. For some other evidence against auxiliary phrase with possum, see Manfredini, in press.

4 In Cormack and Smith, 2002, each negative polarity item is said to be located on a different level: non1 corresponds to pol[neg], or sentential negation, and non2 to vp (verbal phrase) negation. The authors conclude that the English counterparts of possum, can and may, have a complementary distribution respective to each of them: in epistemic reading, may precedes sentence negation (non1), but both can and may in deontic reading locate after sentence negation, but before vp negation (non2).

5 The anonymous reviewers remark both examples (9) and (10) have epistemic reading: my own explanations may well lead to such an interpretation, besides the formal feature of non-agent oriented subjects. But I would like to stress the weight of socially regulated principles, or juridical procedures, in order to understand both cases as expressions of rules rather than of logical inferences.

6 Perception verbs also have experiencer subjects and, according to Magni, 2010, p. 218, they can trigger both epistemic or deontic readings. This can apply to the verb in (14).

7 See Bolkestein, 1980, p. 65, where she speaks of ‘hidden directive’, defined as a directive that “does not name the state of affairs desired nor the behavior to be adopted by the hearer in order to bring it about” (ibid., p. 35).

8 The only counterexample to this observation in my present corpus is Sen. Benef. 3, 19, 2.

9 I reproduce the English translation by John W. Basore, The Loeb Classical Library, London, 1928-1935.

10 For licet as a modal that conveys permission, see, for instance, Orlandini, 1998, p. 1023, n. 11, and Magni, 2010.

11 Note the change in word order, compared to the first instance at the beginning of this passage.

Haut de page

Pour citer cet article

Référence papier

Adriana M. Manfredini, « Auxiliaries in double-negation settings (non potest non + inf): Semantic gliding and syntactic description », Pallas, 102 | 2016, 117-125.

Référence électronique

Adriana M. Manfredini, « Auxiliaries in double-negation settings (non potest non + inf): Semantic gliding and syntactic description », Pallas [En ligne], 102 | 2016, mis en ligne le 20 décembre 2016, consulté le 22 août 2017. URL : http://pallas.revues.org/3618 ; DOI : 10.4000/pallas.3618

Haut de page

Auteur

Adriana M. Manfredini

Chief Assistant, Area of Latin, University of Buenos Aires
adrianammanfredini@gmail.com

Haut de page

Droits d’auteur

Licence Creative Commons
Pallas – Revue d'études antiques est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Presses universitaires du Midi
  • Revues.org