STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world byJSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.istor.org/participate-istor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE ORIGIN OP QUIN-GhAUSES' By Pbank Hamilton Powleb The results of Kienitz" study of quin may be briefly summed up as follows. Quin was a compound of the interrogative adverb qui and the negative ne and so meant originally "why not?"; with this meaning it appears in the imperative-interrogative sen- tence; the use of quin with the imperative and in declarative sentences is derived from the interrogative-negative use; and the conjunction quin, also, is derived from the interrogative adverb. Recent scholarship has in general acquiesced in these results ; but that there are some dissenting voices, especially in respect to the conjunction, reference to the grammars will show.^ Obviously the discussion of the origin of the gm'n-clause should be preceded by a study of the uses of the particle quin in inde- pendent sentences. Whatever theory be held as to the origin of quin in declarative sentences, certainly no one would say that the value actually appearing here is either negative or interrogative. The theory that this use of quin is derived from the interrogative use seen, e. g., in quin taces? goes back, so far as modern scholarship is concerned at least, to the programme of Kienitz mentined above.* This was published in 1878. In 1881 in the second volume of the American Journal of Philology appeared Warren's paper establishing the existence in Latin of a positive non-interrogative particle ne; and in 1888 in the second volume of the Indoger- manische Forschungen, Persson connected this particle with • In the preparation of this paper I have confined myself almost entirely to Plautus and Terence. The complete collections of examples it is impossible to print here, though it is believed that such a presentation would strengthen the argument. * Kienitz De quin particulae apud priscos scriptores latinos usu; Morris Prin- ciples and Methods in Syntax, p. 155 : " In consequence of the abundance and variety of the material, the history of quin is more completely and more surely known than that of any other conjunction ; ' ' Lindsay Syntax of Plautus, p. 109. ' Bennett Grammar 295 3 and 7 ; Hale-Buck Grammar, Index, under "quin." *S'or qui see Kienitz " Qu,i," p. 554. I omit discussion of a possible hercle quin, alioquin, etc. [Classical Philology III, October, 1908] 408 The Oeigin of Quin-Clavses 409 the Indo-European pronominal stem ne, the representatives of which he traced through the several languages. Persson, while admitting the existence of an interrogative-negative quin seen in the interrogative seatences and also, with derived meaning, with the imperative, thought it probable that the quin used with the indicative was a compound of the intensive particle qui, seen in hercle qui, etc., and this positive particle -ne. This derivation has been adopted by Walde in his Etymologisches Worterbuch. I accept this as offering an explanation of the declarative quin perfectly simple and clear and, even if the existence of an inter- rogative quin be admitted, altogether preferable to one in which it is necessary to explain the disappearance of the negative, inter- rogative, and imperative forces supposedly at one time attached to the particle. The meaning of quin declarative is illustrated in the following examples. Most. 456. Eho, an tu tetigisti has aedis? // Cur non tangerem? Quin pultando, inquam, paene confregi foris. // Tetigistin? // Tetigi, inquam, et pultavi Clearly here the quin-sentence assents to the preceding but adds a more important and therefore more emphatic statement. The quin points back to the preceding sentence and indicates a kind of opposition to it as being an insufficient statement. Similarly in Capt. 560. Credidi esse insanum extemplo, ubi te appellavit Tynda- rum. // Quin suom ipse interdxmi ignorat nomen neque scit qui siet. This force of quin is made more explicit by the addition of etiam as in Cos. 93. Quia certumst mihi. Quasi umbra, quoquo tu ibis te semper sequi. Quin edepol etiam, si in crucem vis pergere, Sequi decretumst: So in Men. 806, Mil. 301, 1147, Capt. 289, Cas. 606, Poen. 570. Most. 954 Erras pervorse, pater: Nam nisi hinc hodie emigravit aut heri, certo scio Hie habitare. // Quin sex mensis iam hie nemo habitat. 410 Frank Hamilton Fowlee Here there is opposition to a preceding statement not as being insufficient but as being false. The opposition is less direct but none the less real in the following examples: Merc. 935. Certum exsequist, Operam ut sumam ad pervestigandum ubi sit illaec. // Quin demist. Poen. 672. Rex sum, si ego ilium hodie an med hominem adlexero. // Quin hicquidem tuos est. Persa 485. Credo edepol, credo inqxiam tibi. // lam liberta auctu's? // Enicas. Quin tibi me dico credere. Stick. 593. Edepol te vocem lubenter, si superfiat locus. // Quin tum stans obstrusero aliquid strenue. Quin is never used as a mere responsive; and, on the other hand, it is not used as a mere intensive serving to emphasize its sentence or some word in it. It always points backward and serves to express the relation existing between the preceding sentence and its own, at the beginning of which it stands. That relation is seen to be some kind of opposition varying all the way from "nay, on the contrary," to "yes, and more than that." With this oppo- sition goes very naturally some feeling of impatience or the like. There is no reason why quin with the force which it has in the indicative declarative sentence should not appear also with the subjunctive ; and, on the other hand, there is no special reason why it should. It possesses no force fitting it to assist in expressing a volitive or any other subjunctive idea. If it were true that quin, through its use in the imperative-interrogative sentence, had acquired the power of expressing urgency and so was possessed of a special affinity for the imperative, we should expect to find it showing a special affinity also for the subjunctive in its volitive forces. The fact that it is not so found is evidence that it never had such a force. This rarity of quin with the subjunctive is noted by Hale Cum Constructions, p. 105. In Plautus and Terence the following examples are certain enough to be quoted: Poen. 570. Quin etiam declderint vobis femina in talos velim. Cas. 609. Quin hercle di te perdant postremo quidem. and True. 230, Miles 1125 quoted below. I shall add Mil. 426, by others taken as interrogative. The etiam of the first example The Origin of ^a/iv-CLAusES 411 and the potius of Miles 1125 clearly indicate what is equally true of the other examples, namely that quin with the subjunctive has the semi-conjunctional force seen in its use with the indicative. We may compare the use of immo in Merc. 1015. Eamus intro // Immo dicamus senibus legem censeo Prius quam abeamus, qua se lege teneant contentique sint. Whatever may have been the origin of the quin used with the imperative, certainly that use of the particle furnishes no evidence for either an interrogative or a negative value. It actually pos- sesses neither force ; and whatever may be possible it is not neces- sary to suppose that it was derived from an interrogative -negative adverb meaning "why not?"' So far as I can see the force of quin when used with the imperative does not differ from that of the particle when used with the indicative. Here, too, quin serves to express more or less impatient opposition to something that precedes. The few examples quoted must serve for illustration. PAor. 223. Aufer mi "oportet": quin tu quid faciam impera. And. M9. Quin namst? // Puerilest. // Quid id est? // Nil // Quin die, quid est? Merc. 955. I modo. // Propter istanc. // I modo // Ergocura. // Quintu ergo i modo. In the last two examples the opposition shows more of impatience — impatience at the delay. In the two to follow we have little more than exclamations of impatience: Men. 416. Quin tu tace modo. Rudens 1170. Quid tu i dierecta cum sucula et cum porculis. As in Merc. 955 so in Pseud. 1016, Rudens 628, and Phor. 882 the "pointing-back" force of quin is indicated by the accompany- ing ergo. Of more significance is the quin etiam of Most. 422 and the quin potius of Rudens 1011. There can be no question 1 Ramsay, Excursus to Mosiellaria, regards quin followed by the imperative as elliptical. Quin aspice is equivalent to Aspice, quin aspicis? This, of course, does away with the imperative use of quin. Sonnenschein's insistence that some cases of quin with the imperative should be followed by the interrogation point (C. R. XVI, p. 167 and XXIX, p. 314) I cannot understand. 412 Fkank Hamilton Fowler concerning the propriety of an expression of impatient opposition preceding an imperative or other expression of will, the purpose of the command being to remove the condition to which opposition is felt. The use of quiti with the imperative may be compared with that of immo with the same mood. Most. 583. Quid si hie manebo potius ad meridiem? // Immo abi domum. verum hercle dico: abi domum // Quin vos mihi faenus date. So Merc. 385, Epid. 204, Cist. 521, And. 523, Phor. 935. There are about one hundred and forty gwrn-questions in Plautus and Terence. Only the following need discussion in connection with the statement that quin never inquires for a reason. Pseud. 501. Quin dictumst mihi? is answered by a quia; but the manuscript reading is cur non.^ Stichv^ 516. Quin vocastihominemadcenam? // Nequid adveniens perderem. If the nequid sentence is taken as a purpose clause it would indi- cate a meaning " why not ? " for quin. But the sentence may be an expression of obligation or propriety or it may be an expres- sion of resolve thrown back upon the past (see below p. 422). Merc. 190. Quin, sceleste, earn abstrudebas, ne earn conspiceret pater ? // Quia negotiosi eramus nos nostris negotiis. The emendation of quia to quin is so easy that this single example should have little or no weight in establishing a meaning of "why not?" for quin. Quin does not appear in indirect questions^ with a meaning " why not ? " Contrast the use of qui in Hec. 279. nee qui hoc mi eveniat scio. Of these questions in Plautus and Terence eight have first per- son plural, eight first person singular, and three third person. The great mass, about one hundred and twenty,' have second person. Of the cases with second person all have present tense except four ' Kienitz read quin in Poen. 1317 j but Goetz and Sohoell there read cur non. ^Liv. xl. 47. 4. Cum undique acclamassent, quin ederet, quid fieri velit is simply an imperative question put into indirect discourse. So Livy iii. 62. 14 ; Ourtius v. 5 ; Livy iv. 43. 11. s The ambiguity of -ere of deponent verbs makes exactness here ImposBlble. The Origin of Quin-CijAvses 413 with the perfect and one with the imperfect. There is one case of perfect first person. In the examples with the second person present the act questioned about is one not yet performed but one manifestly desired by the speaker. Hence the questions are vir- tual commands. That they were felt as commands is indicated by the expressions used in referring to them — hortatur in Cas. 764, imperas in Merc. 496, dictum oportuit in Merc. 724, oras in Persa 399, vis in Persa 766, quaeso in Amph. 775 {quaeso with the imperative, e. g., in Asin. 596). The question is paralleled with the imperative in Most. 815, Capt. 636, Trin. 1026, Cure. 611, Pseud. 891, 1183. The questions with the first person plural are also virtually commands. Formally these sentences are questions relative to the act of the person addressed in connection with the person speaking. The first person singular questions are spoken in soliloquy, i. e., the first person is at the same time second person; and the imperative value of the question is manifest here also. Merc. 582. Quid stamusi quin ergo imus .... 1 Rud. 236. Quin voco, ut me audiat, nomine illam suo? Ampelisca. The rare questions with the third person are also imperative in effect : Cure. 251. Palinure, quid stas ? Quin depromuntur mihi Quae opus sunt . . . ? The ecquis questions may be compared — Oap<. 830. Heus ubi estis ? ecquis hie est ? ecquis hoc aperit ostium? Since quin vocasf meant, "Why don't you invite?" or, "Won't you invite?" implying always "It is my will that you invite," we might well expect to find examples like quin vocasti? meaning, "Why didn't you invite?" or "Didn't you invite ?" implying that the act formally questioned about was one which was not accomplished, but one the accomplishment of which was desirable, and hence should have been done.' I quote all the examples of gMzn-questions with a past tense: • In these questions, in the answer to the question in Stichus 576, and in the pas- sage quoted below from Rvdens 378 we come upon the question of the origin of the Subjunctive of Obligation or Propriety. See Elmer Studies in Latin Moods and Tenses; Bennet Critique of Recent Subjunctive Theories; Hale-Buck Grammar. 414 Frank Hamilton Fowler Merc. 189. Bho tu, eho tu quin cavisti, ne earn videret, verbero? 190. Quin, sceleste, earn abstrudebas, ne earn conspiceret pater ? Quia negotiosi eramus nos nostris negotiis. 622. Quin percontatu 's hominis quae fades foret . . . . ? Rud. 841. Quin occidisti extemplo? // Gladius non erat // Caperes aut fustem aut lapidem. Stich.516. Quin vocasti hominem ad cenam? // Nequid adveniena perderem. Trin. 290. lacrumas haec mihi, quom video, eliciunt, Quia ego ad hoc genus hominum duravi. Quin prius me ad pluris penetravi? That which has happened to these questions may happen to a question without quin. Compare with Merc. 189 Rtidens 378. Cavistis ergo tu atque erus ne abiret, quom scibatis f // Quid faceret? // Si amabat, rogas quid faceret? adservaret Dies noctesque: in custodia esset semper. If now the ^wiVt-question never inquires for a reason and is always rhetorical, i. e., always implies a command, it cannot be justly said that we have evidence here for an original meaning "why not?" unless it can be shown that such a question is neces- arily or at least probably derived from one that does inquire for a reason. It is of course admitted that a "why not?" may imply a command. Mil. 1254. Tace, ne audiat. // Quid astitisti obstupida? cur non pultas ? But in view of the numerous sentence-questions' in Plautus and Terence with imperative effect, it can be said with certainty that there is nothing in the nature of tha quin-sentence that demands a rhetorical "why not?" as the meaning for quin. The following may be noted as illustrations: Bacch. 777. Taces? Persa 533. Tacen an non taces ? Cure. 41. Etiam taces? True. 937. Malam rem is et magna magno opere serva tibi viaticum. Phor. 930. In' hinc malam rem cum ista magnificentia? Persa 412. Accipin argentum? accipesis argentum, impudens. Tene sis argentum. etiam tu argentum tenes? Possum te facere ut argentum accipias, lutum? 1 Morris A.J.P. X. 397 ff., XL 16 ff. and 145 fif. The Origin of ^otjv-Clauses 415 True. 373. Dan savium? // immo vel decern. Bacch. 1168. Abin hinc? Aul. 660. Fugin hinc ab oculis? abin an non? Persa 275. Scelerate, etiam respicis? Even if we should admit the literal meaning "why not?" for quin in the pre-Plautine language, it would still be difficult if not impossible to explain the state of affairs in Plautus according to the orthodox view. Would it be possible for a word meaning "why not?" to be restricted in its application to rhetorical use? And how could quin pass over to use with the imperative- — as it is supposed to have done — so long as it retained its interrogative and negative meaning ? If, however, it be said that the conscious- ness of the meaning "why not?" in connection with quin was lost, the last vestige of an argument for such a meaning disappears.' If we place beside such questions as those just quoted the following gttOT-questions, an explanation of the use of quin is at once suggested that is recommended by its simplicity. Merc. 494. Quin taces? // Muto imperas. Men. 915. Quin tu is in malam crucem ? Persa 424.^ Cedo sis mihi argentum, da mihi argentum, impudens. Possum a te exigere argentum ? argentum inquam cedo. Quin tu mihi argentum reddis? nilne te pudet? Cure. 94. Quin das savium? Mil. 1087. Quid hie nunc stas? quin abis? Men. 747.' Quin respicis? That explanation is that quin is used much as in the case of the declarative sentence to give expression to more or less impatient opposition felt to something that precedes. The more direct comparison, however, is with the quin used with the imperative, examples of which have been quoted above. In the first two examples the questions are little more than exclamations of impa- tience. In Persa 424 the quin explicates the impatient opposition felt at the delay in the obeying of a command. In Miles 1087 1 Morris Prin. and Methods, p. 157, remarks that the interrogative force (of quin) is still further weakened by the extension to quin die quin obi; while on p. 134 he has said that the imperative mode is incompatible with interrogation. 2 The importance of this example is increased by the fact that the lines are intended as a kind of echo of 11. 412-414 quoted above. 8 Wagner translates, "Won't you look around?" 416 Frank Hamilton Fowleb the opposition is to the implication of the preceding question, and the same thing is true in the case of Miles 1387, Trin. 802, Haut. 831, Asin. 597. A similar opposition is implicit in such examples of the imperative as Persa 600 Quid stasf adi, Men. 676, Cas. 749, True. 175, Aul. 444. The pointing-back force of quin is sometimes indicated by another particle used with quin or by a conditional or causal clause. For example, potius in Cas. 100, ergo in Merc. 582, Asin. 850, si, ita, etc., in Pseud. 638, quando in Miles 1085. The equivalence of quin in these questions with the quin of declarative sentences is indicated by Cas. 285. Quin, si ita arbitrare, emittis me manu? // Quin id volo. Merc. 724. Quin dicis? // Quin, si liceat- Most. 572. Quin tu istas mittis tricas? // Quin quid vis, cedo. The use of etiam with imperative questions' supports the explanation of quin here advanced. That particle came to have the power of expressing impatient opposition to something that has been said or done. This force appears in the questions of the etiam rogas? type. Amph. 376. Etiam clamas, carnufex ? 381. Etiam muttis? // lamtacebo. Aul. 424:. Etiam rogitas? In such cases the etiam points backward only because the verb to which it belongs refers to that which has just taken place. If now the speaker of Amph. 381 had said, Etiam, tace, it would seem that the circumstances and the tace would make it clear that the etiam was meant to express impatience with the muttering of the other speaker; and of course this would be equally true if some other equivalent form were used in place of tace. My point becomes clearer from a consideration of two interpretations of Phor. 542. Etiam tu hinc abis f The question simply is whether the etiam expresses impatience with something which the abisP is intended to correct or whether the etiam expresses impatience with the act of abis itself.^ There are some twenty-three of these etiam and etiamne imperative 1 Compare Morris A.J.P. X. 431, XI. 41, and XI. 16^. 2 Of. Morris as above and Elmer ad loe. The Oeigin of Qpiw-Clauses 417 questions in Plautus and Terence and five consist of etiam taces? {Cure. 41, Persa 152, Trin. 514, 790, Adel. 550) with which may be compared, taces?, quin taces?, tace, and quin tace. If my explanation is correct such an example as the following would mean, "What, asleep under these circumstances! Won't you wake up?" Most. 383. Etiam vigilas? In this examination of the uses of qiiin we have discovered no evidence of a meaning "why not?" Adopting the derivation of Persson for the quin of the declarative sentence we have found a meaning consistent with that derivation present not only in the declarative sentence but also in the imperative and interrogative sentences. That function consists in the explication of opposition to something that precedes. We come now to the question of the origin of the gMtn-clause. What was the paratactic expression from which that clause has been derived ? And first, is there evi- dence that the quin ever meant "why not?" If the paratactic gwin-sentence was a subjunctive question we should expect to find examples of that independent question actually existing in the literature. The infrequency of this con- struction, Morris thinks, is the sole gap in the history of quin. The only example quoted by anyone with confidence is Mil. 426. Me rogas ? hem, qui sim ? // Quin ego hoc rogem quod nes- ciam. But this is by no means necessarily interrogative; it may be declarative and the subjunctive express the resolve of the speaker.' That the first person singular subjunctive was used in Plautus with volitive force is shown by the examples to be noticed below (p. 421). (Note especially rogem in Trin. 758.) That quin could be used with the subjunctive is shown by the examples quoted on p. 410. That quin could be used with an expression of resolve is shown by Mil. 1124. Quin si voluntate nolet, vi extrudam foras. 1 " Why shouldn't I ask? " is expressed in Stichus 333 by Quidni rogitemf Lucre- tius i. 798 : Quin potius constituas? which Munroe translates " Why not rather hold? " should also be considered declarative, "Nay, rather hold." 418 Frank Hamilton Fowlee Do the clauses themselves show evidence for a meaning "why not?" for quin? In the case of the nulla causa est quin sentences an original interrogative force may be recognized as possible ; but the supposition of such a meaning is not necessary, as I shall attempt to show later. With such expressions as nequeo oontineri, non possum, numquam erit tarn avarus, etc., no interrogative force is discernible. Indeed Morris, though in one place' he speaks of the gwrn-clauses as originally indirect questions, later'' says that the gwm-sentence was not a question but a repudiating exclamation requiring no answer. Obviously there is no evidence here for a meaning "why not?" unless, indeed, it can be shown that the force which quin actually possessed when it became a conjunction was necessarily derived from such a meaning. I follow Kienitz in postulating a single origin for gttm-clauses. This can of course be proved only by tracing the several types from the common source; and this can be done only after that origin has been established. But the fact that quin occurs only after negative principal sentences is very much in favor of the theory of the single origin. In such a construction as licet eat we say that the complex is an expression of permission, consent, etc., and that it is made up of a subjunctive itself expressing permission and a verb licet explicating that force of the subjunctive.' Properly stated the usual explanation of the nulla causa est quin construction is that the quin-question implied that there was no reason against the action, and that nulla causa est explicated that implication. In deterreo ne conferant, deterreo was a prefixed sentence making explicit an implication of ne conferant.* Now whatever may have come to be true later, it certainly was the case when such a para- taxis as deterreo ne conferant arose that the complex as a whole had the meaning of the subjiinctive clause. So in the case of our grwrn-construction we may feel confident that we are on the way to an explanation if we can point out a type of grtm-sentence in which the complex has a meaning proper to the subjunctive and IP. 139. 2 p. 159. ^Compare Morris Principles and Methods, pp. 132 ff. ^Compare Hale-Buck Grammar 503. 3. b. 1. The Origin of Quin-Clavs^s 419 in which the subordinate clause if made independent would have such an implication as is expressed by the principal clause. Our confidence will be strengthened if we can show that the other types of quin-sentence may be derived through extension to other persons, to other tenses, or in other ways. Of the one hundred and sixty examples of gwm-sentences in Plautus and Terence, sixty-six, or 44 per cent, have first person in the quin-c\ause and the sentences as wholes relate to the act of the person speaking. With no sure exception in the first person, only two in the second, and seven in the third, all the sentences relate to an act not yet performed. Of the examples with the first person two-thirds may fairly be said to relate to an act willed by the speaker. As will be shown later all the explicatory sen- tences are easily derivable from an expression of denied hindrance. The following sentences, then, may be taken as fairly exemplify- ing one-fourth the entire number of examples. They represent, I believe, the original type of quin-claxise. Men. 1124. Signa adgnovi: contineri, quin complectar, non queo. Amph. 559. Tamen quin loquar haec, uti facta sunt hie, Numquam uUo modo me potes deterrere. Men. 253. Varum tamen nequeo contineri quin loquar. Audin, Menaechme? Amph.1051. Neque me Juppiter neque di omnes id prohibebunt si volant, Quin sic faciam uti constitui: pargam in aedis nunciam. Miles 1250. Durara nequeo Quin aam intro // Occlusaa sunt foris // Exfringam. Haut. 762. Harcla non possum pati, Quin tibi caput demulcaam: accede hue, Syre: Faciam boni tibi aliquid pro ista ra, ac lubans. Trin. 705. Non enim possum quin exclamem: euge .... Men. 518. Numquam edepol quisquam me exorabit, quin tuae Uxori ram omnem iam, ut sit gesta, eloquar. Rud. 758. Quid causaest qmn virgis te usque ad saturitatem sau- ciem? Quid illas spectas? quas si attigeris, oculos eripiam tibi. Mil. 332. Me homo nemo deterrebit quin ea sit in his aedibus. Hie obsistam, ne imprudenti hue ea sa subrepsit mihi. 420 Frank Hamilton Fowleb Stick. 302. Non enim possum quin revortar, quin loquar, quin edis- sertem Eramque ex maerore eximam, bene facta maiorum meum Exaugeam atque illam augeam insperato opportuno bono. Contundam facta Talthubi contemnamque omnis nun- tios: The obvious meaning of these examples, especially if they are considered in connection with their context, is one of will on the part of the speaker concerning his own action; and the obvious explanation of the origin of such a type is the existence of a sen- tence expressing so much of determination and consequent cer- tainty that the possibility ' of hindrance is rejected and the rejec- tion of such a possibility is made explicit by such an addition as nemo deterrebit or nequeo contineri. The volitive force of these sentences is sometimes indicated specifically. Of the passages quoted above. Men. 253 has an accompanying imperative question; but a clearer indication is to be seen in the accompanying expression of resolve in Mil. 332, Amph. 1051, Mil. 1250, and Stichus 802.' In the following, with the third person to be sure, the volitive force of the gttm-clause is indicated by the parallel ut tie clause. Trin. 105. Est atque non est mihi in manu, Megaronides: Quin dicant, non est : merito ut ne dicant, id est. The illustration of our construction in a paratactic stage is of course impossible. However, the following passages are of interest since they show the collocation of an expression of will with one of negative hindrance. The last two are of greatest value since they have first person. Mil. 1125. Istuc cave faxis. quin potius per gratiam Bonam abeat abs te: Amph. 972. I sane et quantum potes parata fac sint omnia. // Quin venis quando vis introl faxo baud quicquam sit morae. 1 A large number of the cases of negative hindrance have some form of possum and the verb of hindering in the infinitive ; but the possibility of hindrance may be denied as well by numquam deterrebor as by nequeo contineri. 2 It is worth while to note that we may place a full stop after eximam and consider exaugeam and augeam independent subjunctives of resolve. See below. The Oeigin of Quin-Clavs^s 421 True. 230.' Numquam amatoris meretricem oportet causam noscere: Quin ubi nil det, pro infrequente eum mittat militia domum. Merc. 644. Non possum durare: certumst exulatum hinc ire me. Men. 326. lam ergo haec madebunt faxo: nil morabitur. Our theory that the quin-sentence originally expressed the will of the speaker relative to his own act necessitates the supposition that the subjunctive in first person singular was at one time used with volitive force.^ That the most common method of expressing the will of the speaker relative to his own act was in Latin of historical times as in Greek by the use of the future indicative, is well known ; that this could be expressed by the present indicative is also true; but that the subjunctive was at one time in use with this force we should conjecture from its common use to express the will of the speaker relative to the act of another. This con- jecture we find confirmed by the examples, though admittedly few, of the first person subjunctive with this force. The number of examples is, of course, greatly decreased by the ambiguity of the first person singular of the third and fourth conjugations. The following passages are quoted:^ Trin. 1136. Sed maneam etiam, opinor. Phor. 140. Ad precatorem adeam credo. Asin. 605. Sermoni iam finem face tuo: huius sermonem accipiam. Bacch. 1058. Ecfertvir praeda ex Troia. taceam nimciam. Most. 849. Ibo intro igitvir. // Mane sis videam, ne canis — Persa 542. Videam mode Mercimonium. Trin. 758. Dum occasio ei rei reperiatur, interim Ab amico alicimde mutuom argentum rogem. Haut. 273. Mane: hoc quod coepi primum enarrem, Clitipho: Compare- further True. 692, Asin. 816, 610, Cas. 516, Cure. 160, Rudens 570, and other cases which may have future iThis has been explained as declarative, as interrogatiTe, and as dependent. 2 See Delbruck Conj. und Opt., pp. 11 ft.; Vergleichende Syntax II, p. 384; Kie- mann Syntaxe Lat, p. 260; Hale Antic. SubJ., p. 14; Bennett Critique, p. 25; Durham Substantive Clauses, p. 83 ; Morris Prin. and Meth., p. 136. Elmer Studie.s, p. 217, is certainly wrong in explaining as subjunctives of obligation the cases commonly explained as volitive. It may well be that the cases with credo and opinx/r do not show a clear volitive force and yet be volitive in origin, ' In regard to Miles 426 : Quin rogem? see above, p. 417. 422 Frank Hamilton Fowler indicative. I would add the following as showing the extension of the use to the past: Stichus 576. Nequid adveniens perderem. The force of such a subjunctive may be made more explicit by the addition of such expressions as certumst, and so we have: Poen. 501. Profestos testes habeam: decretumst mihi. In the following the form is ambiguous but Morris' and Durham' consider it subjunctive notwithstanding Merc. 472: certumst ibo atque daboj Asin. 248, Amph. 1048, Cas. 448, Bacch. 382, Aul. 681, Capt. 118. With such expressions as those just cited, which are possibly paratactic, Durham connects such substantive clauses as — Amph. 762. ita animatus fui Itaque nunc sum ut ea te patera donem. Compare further Asin. 505, Pseud. 549, Aul. 383, 371, Cure. 218, Men. 1058. In regard to the conjunction quin it need hardly be said that the meaning which we have found attached to the particle in all its other uses is one that exactly fits it to show the relation between a loquar, "I shall speak," and the added explication, nequeo contineri, "I can't be stopped." Nequeo contineri: quin loquar originally meant, "I can't be stopped: nay, I will speak." An objection to the theory of the interrogative origin of the ^Mm-clauses is that quin is not found with the subjunctive in questions (see p. 417) . That quin is not found commonly with the subjunctive of resolve does not militate against the theory now proposed. As we have just seen, that subjunctive was itself rare in Plautus' time; the one example which I claim above [Mil. 426: quin rogem) is as much as we should expect. And quin was used with other expressions of resolve, e. g., Mil. 1124, quoted above. In what follows I shall not attempt to trace the development of the gwm-construction in detail. My object is simply to show, as an argument in support of the theory advanced in this paper, that 1^. J. P. XVIII, p. 145, and Prin. and Meth., p. 136. ^Subst. Clauses, p. 84. The Oeigin of Quin-Glavs^s 423 the several types of quin-c\a\is& which we find in Plautus and Terence are only those which we might expect to find if we assume as the original type that illustrated by nequeo contineri quin loquar. This sentence is a complex, as a complex in first person, with volitive modal force, with future temporal force, and with a, gram- matically speaking, principal clause expressing negative hindrance. Any one of these elements may show shifting, but the modal force not commonly, except when accompanied by shifting of some other element. It is to be noted, too, that in our typical sentence the loquar expresses in effect that which does not follow as a result of the nequeo contineri. This relation of virtual negative result may become more prominent in the developed clause. The extension to the past is not common. For the first person the references are as follows: Mil. 370, Asin. 675, Eun. 1043, Bacch. 1012, Amph. 1054, Adel. 222. Of these Mil. 370: Numquam hercle deterrebor Quin viderim id quod viderim is only a formal extension. As Brix notes, the expression is short for vidisse me dicam. In Asin. 675 the time of the quin-verh is really future, and the imperfect is due to the imperfect subjunctive of mental certainty upon which the quin-clanse depends. Only in Adelphoe 222 do we have both verbs in a past tense. The occur- rences in the second person are JSun. 180 and Haul. 1007 ; those in third are Eun. 842, Trin. 534, Cist. 18, Eun. 1092, Cure. 228, and Miles 263. It may be worth while to note that of these fourteen cases six are from Terence, the later writer. Of more interest is it to note that most of these extensions have taken place only when the explicatory sentence had gone some distance from the starting-point. Most of the cases are like Haut. 1007. nuUamne ego rem umquam in vita mea volui quin tu in ea re mi fueris advorsatrix, Sostrata? But in the following we have a true extension to past tense and to third person without other shift. Cure. 228. Tormento non retineri potuit ferreo, Quin reciperet se hue esum ad praesepem suam. Similarly, Mil. 263. 424 Fbank Hamilton Fowlee The number of cases of gMtre-clause with the third person is the same as of those with the first/ In the following the modal feeling is volitive. Merc. 1021. Neu quisquam posthac prohibeto adulescentem filium Quin amet et scortum ducat, quod bono flat mode. Adel. 170. Cave nunciam oculos a meis oculis quoquam demoveas tuos Ne mora sit, si innuerim, quin pugnus continuo in mala haereat. But such an example as the following expresses consent. Phor. 272. Non causam dico quin quod meritus sit ferat. Of quin-c\auses with the second person there are but twenty-two cases. The modal feeling is usually shifted. In the following it is consent. Cure. 34. Nemo hinc prohibet nee vetat Quin quod palamst venale, si argentumst, emas. Asin. 355. argentum non morabor quin feras. We may expect the gwm-clause to be extended to use with other less direct expressions of negative hindrance ; and this shifting we may expect to find accompanied by a shifting of modal force and by a change in the relation existing between the two clauses. The use of non sinam quin "I shall not permit a hindrance," non patiar, nequeo durare quin, "I cannot hinder myself," scarcely call for notice. See example quoted above, p. 419. These ex- pressions do not occur except in first person. And there is no more difficulty with non possum and nequeo quin, though the feeling here may be weakened from the true volitive. See Trin. 705, quoted above. Since the nulla causa est quin construction^ has been thought by some to furnish the starting-point for all gwm-constructions, and by others to be at least independent in its origin, it calls for more attention. There are really three types of these construc- tions, the nulla causa est quin with five examples in Plautus and 1 Such an example as the following, while having the verb in the third person is practically first person : Miles 332. Me homo nemo deterrebit quin ea sit in his aedibus. 2 Mr. Tenney Frank's article dealing with these sentences appeared in the January number of this Journal after my paper was in the hands of the editor. The Origin of Quin-GijAvsbs 425 Terence, the nullam causam dico quin with four examples, and the quae causa est quiw with, seven examples. One example, Amph. 852, belongs in both of the last two classes. In such an example as Cas. 1003. Nulla causa est quin pendentem me, uxor, virgis verberes. the existence of a hindering reason to another's action is denied and modal feeling becomes one of consent. As regards the modal feeling the same is true of Aul. 755. Ergo quia sum tangere ausus, baud causificor, quin eam Ego habeam potissimum. The speaker denies that he will offer a pretext that will stand in the way of the act consented to.' With more of the original feel- ing of resolve are such cases as Rudens 758, quoted above. Here the speaker challenges — and so virtually denies — the existence of a hindering reason for the action resolved upon. How close the nulla causa construction is to that after an expression of negative hindrance is shown by such an example as Hec. 588. Ne mea praesentia obstet neu causa ulla restet relicua, Quin tua Philumena ad te redeat. The few cases of non facio quin are close to the nulla causa est quin construction. The close relation between the nequeo durare and the non faciam type and through that with the nulla causa type is shown by Amph. 889. Non edepol faciam neque me perpetiar probri Falso insimulatam, quin ego ilium aut deseram Aut satis faciat mihi ills atque adiuret insuper Nolle esse dicta quae in me insontem protulit. In place of denying the existence of a hindrance in general we may have the denial of a specific hindrance. Cas. 504. Tribus non conduct possum libertatibus Quin ego illis hodie comparem magnum malum Quinque hanc omnem rem meae erae iam faciam palam. Men. 518. Numquam edepol quisquam me exorabit, quin tuae Uxori rem omnem iam, ut sit gesta, eloquar. Most. 146. non videor mihi Sarcire posse aedes meas, quin totae perpetuae ruant ' Cf . Elmer's note to Phor. 272 " following the idea of preventing implied in non causam dico." 426 Feank Hamilton Fowler The first and second examples are clearly volitive in force. In the last with the third person the consecutive relation becomes more prominent. The degree of a quality may be made the specific hindering reason denied. Capt. i08. Numquam erit tam avarus, quin te gratus emittat mauu. Haut. 675. Nil tam diflficilest quia quaerendo investigari possiet. And here, too, the consecutive force is prominent.' It is only a step from such a case as that just quoted or Haut. 805. NuUast tam facilis res, quin difficilis siet, Quam invitus facias to such a one as Stick. 208. Nam curiosus nemost quin sit malevolus — : and from this only a step to such a one as Phor. 697. Nil est, Antipho, Quin male narrando possit depravarier. Brugmann I. F. IV. 226, thought it impossible than an adverbial clause should pass over into an adjectival clause and so separated these g;Mm-clauses from others. But, as already indicated by the examples quoted from Heauton and Stichus, these were not felt as adjectival clauses. More conclusive evidence is furnished by the passage quoted above from Haut. 1007 in which the demonstrative in ea re shows that quin was not felt as a relative pronoun. The same kind of evidence is furnished by the following examples: Trin. 534. Neque umquam quisquamst quoius ille ager fuit, Quin pessume ei res vorterit. Nepos xviii. 2. 5. Non cum quoquam arma contuli quin is mihi suc- cubuerit. In no case, however, is it to be admitted that we have a true result clause. Quin never became grammatically equivalent to ut non. To say that nothing is so easy as to prevent its being difficult under given circumstances may be practically equivalent to saying that nothing is so easy that it is not difficult under given circum- stances ; but grammatically the two expressions are not equivalent. 'Cf. Bennett Appendix, 376. The Origin of Qum-ChAvs^s 427 Just as in the case of the true negative result clause the mean- ing is sometimes that which we may translate into English by "without" and the gerund or into German by "ohne dass," so in such cases as the following, Epid. 437. Cave praeterbitas ullas aedis, quin roges. With the nemo est quin construction should be compared, e. g., Adel. 294. nam numquam unum Intermittit diem quin semper veniat. With nil tarn est quin compare Haut. 68. Numquam tam mane egredior neque tam vesper! domum revortor quin te in fundo conspicer fodere aut arare aut aliquid ferre. It is worth while to note that there are only nine cases of the non dubium quin construction in Terence and only three in Plautus. These three have quid dubitas? The construction was somewhat figurative and grew up later than the others. The explanation of the construction lies in the fact that dubium means "hindrance to thought.'" And. 172. Non dubiumst, quin uxorem nolit Alius. Lombard College ' Compare Mil. 332 quoted above, p. 419. With other meanings of dvbium and dvbito the use of quin iS rare and comparatively late.