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 TEMPORAL CC/^ilf-CLAUSE AND ITS RIVALS By J. J. SCHLICHBE The various forms of the cwm-clause perform approximately the same temporal functions as the clauses introduced by post- quam, ubi, ut, simulatque, and dum. The two sets of construc- tions are, in a sense, interchangeable. Sallust, in his Catiline and Jugurtha, has 209 cases of postquam and ubi, and 24 cases of temporal cum with the subjunctive; in Caesar the figures are 77 and 147, and in Nepos 49 and 181, respectively. There is no reason to doubt that if Sallust had written the Commentaries or the Biographies, he would still have used the conjunctions in sub- stantially the same proportion as he did. His individual prefer- ence led him to use postquam and ubi, which shows that while these may be regarded as substitutes for cum, there is neverthe- less a difference between them and cum, which determined his choice. And this difference is clearly not an objective one, due to the character of the facts narrated, but subjective, reflecting the mental processes of the writer. To discover it we must, there- fore, examine the mental processes involved in such forms of speech, rather than the external facts which they express. The thinking which accompanies our speech is progressive.' We do not have the whole subject before our minds with equal clearness at any moment, while we are speaking or writing, nor even so much as is contained in a single long sentence. Rather, the different things which we say move, one after the other, into the full light of our consciousness, where they are held long enough to be seen in their details and internal structure, preparatory to their expression in speech. Then they return to the vague back- ground of our minds, from which they came. 'See Wundt Volkerpsychologie — Die Sprache, Vol. I. 23i-48, and Morris On Prin- ciples and Methods in Syntax, chap. ii. [Classical Philology IV, July, 1909] 256 The Temporal Cpm-Clause and its Rivals 257 It is evident that for any given section of thought there is what we may call its culmination, the stage when it has attained its maximum of clearness in the mind. This must necessarily be so, since it passes from a state of vagueness, before the moment of speech has arrived, through one of distinctness into another state of vagueness after its expression, when the speaker's attention has passed on to something else. If this is true, it is also clear that when the vocal organs are ready to begin a new clause, the idea contained in it has not in all cases reached the same stage of distinctness, or, as we might say, it is not in all cases equally far advanced toward maturity. So we may, for the purposes of our present discussion, divide the thoughts to be expressed by temporal clauses into two classes — those which are fully matured when expression begins, and those which are not fully matured. In the former the full bearing of what is to be said, and its internal structure, is clearly evident to the mind, while in the latter it is not yet thus evident, since the culmination in the progress of the thought has not been reached. Whether it is the one or the other ought to appear, if a distinction is made in speech at all, in the character of the introductory word. For when the thought is mature and has, even at this point, fully taken shape, the speaker is in a position to be more specific and accurate in his manner of expression than when this is not the case. We shall assume then, as an initial step, that the clauses intro- duced by postquam, ubi, ut, simulatque, and dum probably express thought in a more advanced stage of formation than those intro- duced by cum. For while the words in the former list are all specific in their meaning, cum may be used to introduce a clause when neither the exact temporal relation, nor the nature of the action (aoristic or progressive), nor the nature of the relation itself (temporal, causal, adversative) is as yet distinctly realized by the speaker.^ The tendency of the clauses with postquam, ubi, etc., to con- fine themselves to a single form of the verb, as compared with the wide variety of tense and mood following cum, points in the same 'The process of clarification in the mind continues while the clause is being expressed, and the verb at the end regularly supplies some of the deficiencies of the introductory word. 258 J. J. SOHLIOHEB direction. On the one hand, the specific character of the intro- ductory word in the former is naturally associated with specific forms of the verb. Simulatque requires a verb of momentary action; dum, "while," a tense of continuance; ubi and ut, ordinarily expressing identity of place and manner, and dum, "so long as," naturally make for a form of the verb similar in force to that of the main clause. On the other hand, when the introductory words are themselves specific and indicate fully the temporal relation of their clause to the rest of the sentence, there is little left for the verbs to do. As a result, the verbs with postquam, etc., do little to express this relation. In most cases, with the exception of dum, these con- junctions introduce an act or occurrence preceding that of the main clause. And yet, the present perfect and pluperfect are rarely used with them. And dum, "while," has a form of the verb which does not make any pretense whatever of expressing the actual time of the occurrence which it mentions. With cum, on the other hand, both the character of the action and its temporal relation to that of the main clause, is regularly expressed by the form of the verb, and nbt by the conjunction. II How does it come about that some thoughts reach a stage of maturity in the mind so much more rapidly than others? The easiest answer is that it depends very much on the individual and the circumstances. The minds of all persons do not work with the same speed. Temporary excitement or unusual interest may have an effect. Familiarity with the matter under discussion, the speaker's purpose, the literary fashions of his time, may all play a part. But after this allowance is made, there will still be a suf- ficient residuum of similar conditions in thought, with similar effects in speech, from which reliable conclusions may be drawn. When we speak, one idea follows the other because it is in some way associated with it. The connection between them may be of many different kinds, and may vary greatly in the force and fulness with which the first suggests the second. It is not always the path of least resistance that is followed. For the direction in The Tempoeal Ccrjif-OLAUSE and its Rivals 259 which the thought proceeds will, at every step, be determined necessarily by the underlying purpose and nature of the discourse. And this will call for difficult as well as easy transitions. When the transition is easy, we shall naturally expect the new idea, other things being equal, to make more rapid progress than when it is hard. Generally speaking, the transition should be easy in all those cases where the following idea involves a sequel to the preceding one which is natural or inevitable, and so can easily be foreseen. The transition will be more difficult, and the full maturing of the new idea in the mind will consequently be relatively more slow whenever it involves an occurrence or condition of things which is less closely and necessarily related to what has gone before. As examples of easy transition may be mentioned such clauses as express the arrival at a goal, or the consummation or completion of an action which has been mentioned previously, or the reaching of a certain stage naturally implied in it, or a sequel otherwise directly suggested by it. Of course the mere repetition of a pre- vious idea, with or without elaboration, would also belong here. Clauses introduced by postquam, ubi, ut, and simulatque are widely used in all these cases. Arrival at a goal : Hoc nuntio commotus, sperans se etiamtum pecunia et potentia instans periculum posse depellere, domum rediit. Hue ut venit, ab ephoris in vincla publica est coniectus. (Nepos Paus. 3. 5.) Cum .... eos in eo loco, quo tum assent, suum adventum exspectare iussisset, paruerunt. Eo postquam Caesar pervenit, obsides, arma, servos .... poposcit. (Caesar B. G.i.21. 3.) Itaque paucis diebus cum auro et argento multo Romam legates misit Sed ubi Romam legati venere .... tanta commutatio incessit ut (Sallust Jug. 13. 7.) Itaque, ut domum Cybebes tetigere lassulae, Nimio e labore somnum capiunt sine Cerere. (Catullus Ixiii. 35, 36) Consummation of an action : Jugurtha quam maximas potest copias armat .... Adherbal, tam- etsi Romam legates miserat .... fretus multitudine militum parabat armis contendere. Sed ubi res ad certamen venit, victus ex proelio pro- fugit. (Sallust Jug. 13. 4.) 260 J. J. SOHLIOHEE Is de honore regni cum Agesilao, patruo suo, contendit neque id quod petivit consecutus est. Nam Lysandro suffragante Agesilaus antelatus est. Hie simulatque imperii potitus est, persuasit Lacedae- moniis ut (Nepos Ages. 2. 1.) Tum victu revocant vires fusique per herbam Implentur veteris Bacchi pinguisque ferinae. Postquam exempta fames epulis mensaeque remotae, Amissos longo socios sermone requirunt. (Virgil Aen. i. 214 ff.) Reaching a certain stage in the action : Huic in respondendo Epaminondas, cum de ceteris perorasset, post- quam ad ilia duo opprobria pervenit, admirari se dixit stultitiam rhetoris Attici. (Nepos Epam. 6. 3.) Ubi vero ea pars turris, quae erat perfecta, tecta atque munita est ab omni ictu hostium, pluteos ad alia opera adduxerunt. (Caesar Bel. Civ. ii. 9. 5.) Postquam prima quies epulis mensaeque remotae, Crateres magnos statuunt et vina coronant. (Virgil Aen. i. 723, 724.) Postremo ut iam agendae rei tempus visum est, stipatus agmine armatorum in forum inrupit. (Livy i. 47. 8.) Simul primo concursu concrepuere arma, hostis dolo non metu pedem rettulit. (Livy vi. 24. 1.) Sequel directly suggested: Quaerit quibus locis sit Aspis: cognoscit haud longe abesse pro- fectum que tum venatum. Quem dum speculatur, adventus eius causa cognoscitur. (Nepos Datam. 4. 4.) Micipsa paucis post diebus moritur. Postquam, illi more regio iusta magnifice fecerant, reguli in unum convenerunt. (Sallust Jug. 11. 2.) Pompeius interclusus Dyrrhachio, ubi propositum tenere non potuit, secundo usus consilio edito loco .... castra communit. (Caesar Bel. Civ. iii. 42. 1.) Ita renovatum foedus, indictumque iunioribus Latinorum ut ex f oedere die certa ad lucum Ferentinae armati frequentes adessent. Qui ubi ad edictum Komani regis ex omnibus populis convenere .... miscuit manipulos. (Livy i. 52. 6.) Repetition of the preceding idea : Postquam, ut dixi, senatus in Catonis sententiam discessit, consul .... tresviros quae ad supplicium postulabat parare iubet. (Sallust Cat. 55. 1.) Haec dum Dardanio Aeneae miranda videntur. (Virgil Aen. i. 494.) The Temporal Ocjm-Clause and its Kivals 261 Dum ea Romae geruntur. (Salltist Cat. 56. 1.) Id ubi dixisset. (Livy i. 32. 13.) Haec ubi dicta. (Virgil Aen. i. 81.) Koseis ut huic labellis sonitus abiit celer. (Catullus Ixiii. 74.) All these classes of ideas are also expressed by cwm-clauses, with the difference, however, that they make up a much smaller proportion of the total number of temporal clauses with cum, than they do in the case of the other clauses mentioned above. The following examples will illustrate. Arrival at a goal : lamque ibat dicto parens et dona Cupido Regia portabat Tyriis, duce laetus Achate. Cum venit, aulaeis iam se regina superbis Aurea composuit sponda mediamque locavit. (Virgil Aen. i. 695 ff.) Consummation of an action : Imperavit quam plurimas venenatas serpentes vivas colligi easque in vasa fictilia conici. Harum cum effecisset magnam multitudinem. (Nepos Han. 10. 5.) Reaching a certain stage in the action : Sedutis Hannibal certaminibus, cum iam Alpes peteret (Livy xxi. 31. 9.) Sequel directly suggested : Indutiomari Treveri nuntiis impulsi suos concitaverunt subitoque .... ad castra oppugnatum venerunt. Cum celeriter nostri arma cepissent vallumque ascendissent (Caesar Bel. Gal. v. 26. 2.) Repetition of the preceding idea : Id quantae saluti f uerit universae Graeciae, bello cognitum est Persico. Nam cum Xerxes et mari et terra bellum universae inferret Europae (Nepos Them. 2. 4.) The use of cum when it is possible to use a more specific con- junction must be set down largely as a matter of taste.' From its prevailing association with a certain type of idea, a conjunction with its attending clause came to have a specific atmosphere, and acquire the capacity for producing a specific effect, which the writer of artistic instincts would not be slow to use for his purposes. 1 There are clauses with cum, of course, for which none of the other clauses would be an exact substitute. 262 J. J. SOHLIOHEE At the same time it should be borne in mind that it is much more often true that cum encroaches upon the province of the other conjunctions than that they encroach upon it. This is entirely natural. For it is easy to use a less specific word or clause than the state of your thought will warrant, but not so easy to be more specific in your language than in your thought. Cum is the tem- poral factotum, which can be readily turned to any kind of work, while the other conjunctions are specialized in their functions, and rather useless in any line but their own. But there are, besides this, two additional factors which must not be overlooked, and which may frequently make the use of cum the natural form of expression, even for the various classes of ideas above enumerated, which ordinarily would seem to call for the other conjunctions. If we are right in assuming that the clause with cum is the natural form of expression for the idea which at the beginning of the clause has not yet reached the stage of entire clearness in the mind, it would also seem natural to suppose that the same con- struction might again be used when the idea has already passed its culminating point, and is on its way back to the vague state from which it came. But this is, unfortunately, a stage in the idea's progress where it is impossible to draw a line with any certainty. The upward movement of the idea, from vagueness to clearness, can in a degree be measured, because the goal is in a way fixed, but its decline back into vagueness lacks this definite limit, and hence its progress in this direction escapes our control. We shall therefore have to be content with merely noting the probability that cum may be used for this purpose. Clauses which readily suggest themselves as belonging here, are those which refer once more, very briefly, to the same thought that engaged the speaker's attention more fully and at greater length in the preceding sentence or sentences, and whose culmina- tion clearly lies there and not in the brief second reference to it in the new clause. A common case is the one where a preceding speech is referred to by "haec cum dixisset," or some similar expression. Such clauses are nothing more than an echo of what has preceded, a mere transitional element or conjunction, by means The Temporal Octm-Clause and its Rivals 263 of which the speaker passes on to what follows. They tend, there- fore, to become stereotyped and, to that extent, to be removed from influences which might have an effect on their form. As a result, we find a tendency in the individual author to confine himself to a single form of this expression. So Caesar and Nepos always introduce expressions like the one just mentioned by cumj Sallust, who uses cum but little, prefers postquam; and Virgil re- peatedly uses the phrases haec ubi dicta and haec ubi dicta dedit} In addition to those ideas which have already passed their zenith when the utterance of the clause begins, there are others which never reach full maturity in the mind at all. They may either in themselves not be of sufficient consequence to the speaker, whether absolutely or relatively, to enlist his full effort and attention, or it may seem to him that they are so self-evident, so completely to be taken for granted, that he makes a display of slighting them, and leaving them by the way, as it were. A com- parative vagueness and immaturity of the idea may thus be due to a lack of interest in it, either real or assumed, on the speaker's part, and so to a lowering of the tension in his mind, as compared with that which accompanies the other ideas before and after it which are felt to be of greater importance. It is evident, of course, that in the case of such an idea, which is slighted by the speaker, or treated as requiring little attention, the character of the action involved and its relation to other actions will be less clearly marked in his mind than would be the case if his interest and attention were fully enlisted. That is to say, the mental status of such an idea is similar in this respect to that of an idea but partially matured, or one whose point of culmination is already past. The clauses introduced by cum in the following sentences will illustrate what is meant: Vidi etiam senem Livium qui, cum sex annis antequam ego natus sum fabulam docuisset Centone Tuditanoque consulibus, usque ad adulescen- tiam meam processit aetata. (Cicero Sen. 14. 50.) 1 Perhaps no form of temporal clause becomes stereotyped more easily than dmn with the present indicative. A very large proportion, in some authors, like Caesar, more than haU, of all the occurrences of this conjunction, are in clauses with the verb geruntur (dum haec geruntur, and the Uke), referring to an action just described. The stereotyped character of this clause is, of course, also apparent from the use of the present tense. 264 J. J. SCHLICHEE Haec loca vi quondam et vasta convolsa ruina — Tantum aevi longinqua valet mutare vetustas — Dissiluisse ferunt, cum protinus utraque terra Una foret. (Virgil Aen. iii. 414 fif.) In secundo proelio cecidit Critias, dux tyrannorum, cum quidem adversus Thrasybulum fortissime pugnaret. (Nepos Thras. 2. 7.) Gaius Valerius Procillus, cum a custodibus in fuga trinis catenis vinctus traheretur, in ipsum Caesarem hostis equitatu persequentem incidit. (Caesar B. O. i. 53.) In rapid narration we frequently find ideas slighted in this way in some authors, which by others are regularly given full prominence. Thus the arrival at a goal which has been previ- ously mentioned is in Caesar expressed much more frequently by "eo cum venisset" and the like, than by clauses with uhi and postquam, which might be called the more normal method of expressing such an idea. The explanation is, no doubt, that in rapid narration, like that of Caesar, where the thing to be accom- plished is the matter of chief concern, rather than the details of the process, the author does not care to dwell a second time with full attention upon a goal, let us say, which, it is evident from the preceding sentence, will surely be reached. On the other hand, a striking proof of the eifect produced upon the character of the temporal clause by excitement and intensity of mental action is found in two of the poems of Catullus. The "Marriage of Peleus and Thetis" has cum twelve times, simul{ac) six times, postquam four times, and dum and ut once each, which is about the proportion that we find in epic poetry generally. The " Attis," on the other hand, has ut four times, ubi and simul{ac) each three times, and dum once, but contains no cum whatever. Generally speaking, the poets use cum, especially circumstantial cum, much less than the prose writers. There are, however, exceptions among the latter, like Sallust and Tacitus. That vigor and clearness of thought and the possession of the speaker's full attention are factors favorable to the specific tem- poral clauses is further shown by what we find to be true in the expression of certain other ideas, which not only have the advan- tage of rapid maturity, but in addition to it are so situated that, The Tempobal Ctiiif-CLAUSE and its Kivals 265 for special reasons, they command the attention in an unusual degree. We find this state of things in cases like the following: Quae (vasa fictilia) iacta initio risum pugnantibus concitarunt, neque quare id fieret poterat intellegi. postquam autem naves suas oppletas conspexerunt serpentibus, nova re perterriti .... puppes verterunt. (Nepos Han. 11. 6.) Vagenses .... primo, uti res erat, Metellum esse rati, portas clausere, deinde ubi neque agros vastarl et eos qui primi aderant Numidas equites vident, rursus Jugurtham arbitrati cum magno gaudio obvii procedunt. (Sallust Jug. 69. 1.) Eegem primo delectatum esse munere: post ubi audisset unde essent, statim certos homines in quinqueremi misisse. (Cicero Ver. iv. 103.) Namque antea pleraque nobilitas invidia aestuabat, et quasi poUui consulatum credebant, si eum quamvis egregius homo novus adeptus foret. Sed ubi periculum advenit, invidia atque superbia post fuere. (Sallust Cat. 23.) It will be noticed that here the idea expressed by the clause with postquam and ubi is anticipated and prepared for in what precedes [initio, primo, antea, deinde, sed),' and that it has, therefore, a comparatively long time in which to mature. Con- sequently, it may reasonably be assumed to be in shape for clear and definite expression when its turn comes. On the other hand, interest in it is assured by the fact that it stands in contrast to the statement before it, and concentration of the attention upon it is thus secured. So it comes that we find this sort of temporal clause introduced by one of the definite conjunctions almost with- out exception. The following figures will show, for the three authors mentioned, the number of such contrasted clauses among the total of those introduced by postquam, ubi, and ut, and of the temporal clauses introduced by cum. Postquam, etc. / Cum Caesar 12 out of 82 14 out of 75 39 out of 209 4 out of 1.57 Nepos 3 out of 194 Sallust 1 out of 46 ' Of course, expressions suggesting the contrast are not always thus fully employed. Sometimes the contrast is implied in the construction of the whole passage, as in Nepos Ale. 11. 3 and 1. 4. 266 J. J. SOHLIOHEK A still further class of expressions which have the combined advantages of rapid maturity and mental tension or interest, is found in cases where a person, occurrence, or state of things, pre- viously mentioned or described, is represented as coming to the notice of someone who is prominent in the action of the passage, and as leading to a change of course or the adoption of a new line of action on his part. Clauses of this kind are numerous in all kinds of narrative. lUe ubi accepit homines claros .... contra inceptum suum venisse. (Sallust Jug. 25. 6.) His in ignem coniectis flammae vim transiit. Quem ut barbari incendium effugisse viderunt .... (Nepos Ale. 10. 6.) Postquam id animadvertit, copias suas Caesar in proximum collem subduxit. (Caesar B. G. i. 24. 1.) Haec simulac cupido conspexit lumine virgo. (Catullus Ixiv. 86.) The occurrences, persons, and things referred to are, of course, clearly present in the speaker's mind at the beginning of the tem- poral clause, since they have just been spoken of. The idea of their coming to the interested person's notice is an easy one to grasp. That is, the whole thought contained in the clauses beginning with postquam, etc., is such that it may be supposed to mature quickly in the speaker's mind. And the fact that a new turn is given to the course of things by someone's noticing what has happened and acting upon it, furnishes a stimulus to the attention. Naturally, therefore, these clauses also are very much more frequently introduced by ubi, ut, postquam, or simul- atque than by cum. In the table below the figures in parentheses give the total number of occurrences of the conjunctions in each author, those before the parentheses give the number of clauses of the kind we are describing. Caesar Nepos Sallust Virgil Aen. i-vi Ovid Met. i-v J^ofiiouatyi* 13(22) 35(55) 2(5) 4(6) 14(45) 2(4) 12(26) 1(4) 23(90) 30(119) 2(19) 3(30) 10(22) 1(1) 9(30) Ubi 4(25) jjt 12(36) fti7nulo,ta'ue. 7(13) 1(10) 10(147) 0C13) 12(181) 0(22) 1(24) 0(51) 0(11) 0(35) dum (w. subi,^ 1(13) *Tbe clauses introduced hy posteaquam are also included. The Tempobal Ocraf-CLAUSE and its Rivals 267 It appears then that of 552 temporal clauses introduced by definite conjunctions in these authors, 182 belong to the class described above, while out of 507 temporal clauses introduced by cum only 25 belong to that class. Ill Our discussion has thus far dealt at some length with temporal clauses to which the transition is easy. We shall now turn to those clauses to which the transition in thought is more difficult. They will be such as contain ideas that are not so directly in line with what has gone before, and these are, consequently, not anticipated and prepared for in the same way as the ones of which we have spoken heretofore. They reach the mind as matter com- paratively new, and will, other things being equal, require more time for their full maturing. We find that the form of expression gives evidence of this fact in a very decisive way. For the clauses in question are, generally speaking, and with few exceptions, introduced by cum. And clauses of this kind, which contain ideas that are not directly suggested in what precedes, do, moreover, form a large proportion, something like a third or a fourth, of all the temporal clauses introduced by cum.^ Clauses introduced by the definite conjunctions, on the other hand, are, for the most part, rarely used for ideas of this kind. In Nepos, for example, there appear to be only two clauses which are certainly of this kind, among 79 introduced by postquam, ubi, ut, and simulatque, and these are both found in a passage (Dion 4. 3-5) where the author has run ahead of his story, a fact which he realizes himself, as is shown by the words sed illuc revertor at the end of it, and by his immediate resumption of the narrative at an earlier point. In Caesar there are six such clauses, all intro- duced by ubi. But the idea which they express, while not directly anticipated in the preceding sentence, is yet involved in it as one of only two possibilities. They are all cases like the following:" 'No sharp line can be drawn, of course, between these clauses and those whose ideas are directly suggested by what precedes. The one class shades off gradually into the other by many intermediate stages. 'The others are B. G. iv. 21. 3 ; ii. 8. 3 ; yii. 26. 4 and B. C. i. 84. 2 and iii. 41. 2. 268 J. J. SOHLICHBB Proximo die instituto suo Caesar e castris utrisque copias suas eduxit paulumque a majoribus castris progressus aciem instruxit, hostibus pug- nandi potestatem fecit. Ubi ne turn quidem eos prodire intellexit circiter meridiem exercitum in castra reduxit. (B. G. i. 50. 1-2.) The idea expressed by the temporal clause in such a case does not, of course, present a difficult transition in the same degree as the ideas in clauses like those given below, which are regularly introduced by cum: Is pagus appellatur Tigurinus : nam omnis civitas Helvetia in quattuor pagos divisa est. Hie pagus unus, cum domo exisset patrum nostrorum memoria, L. Cassium consulem interfecerat et eius exercitum sub iugum miserat. (Caesar Bel. Gal. i. 12. 5.) Equites nostri cum funditoribus sagitariisque flumen transgress! cum hostium equitatu proelium commiserunt. Cum se illi identidem in silvas ad suos reciperent ac rursus ex silva in nostras impetum facerent .... interim legiones sex, quae primae venerant, opere dimenso castra munire coeperunt. (Caesar Bel. Gal. ii. 19. 5.) Fuga salutem petiit, quam consecutus non esset, nisi intra sua prae- sidia se recepisset, quae in proximo litore erant collocata. Keliquae Pergamenae naves cum adversarios premerent acrius, repente in eas vasa fictilia conici coepta sunt. (Nepos Han. 11. 5.) Regni certamine ambigebant fratres; maior et qui prius imperitarat, .... minore ab fratre et coetu iuniorurn .... pellebatur. Huius sedi- tionis peropportuna disceptatio cum ad Hannibalem reiecta esset, arbiter regni factus imperium maiori restituit. (Livy xxi. 31. 7.) The striking thing about the matter is that in an author like Sallust who introduces such a large number of his temporal clauses by definite conjunctions, to the exclusion of cum, we still find the distinction between the respective provinces of the two classes of expression scrupulously observed. Little as he uses cum, it intro- duces clauses that contain new matter in about the usual proportion of cases, while, on the other hand, of 209 clauses introduced by postquam and ubi, only about eight of this kind are found.' These figures are certainly full of significance from any point of view that we may adopt. If we assume an antecedent preference on Sallust's part for ubi and postquam, we are led to conclude that in order to indulge this preference, he was willing to curtail iJug. 93. 3; 70. 6; 11. 9; 25. 10; 64. 1; Cat. 11, 39, 28. This does not include the clauses expressing repeated action, which will be discussed later on. The Temporal Cujf-CLAUSB and its Rivals 269 considerably the freedom of movement in his thought. And if we assume that his thinking was by nature thus clear and definite, avoiding all half-lights and vagueness, it will follow that he found its proper expression in the clauses with postquam and ubi, which he uses so abundantly. IV Being thus confirmed in our view that the closeness of the relation between a temporal clause and what precedes it is of great importance in determining the kind of clause to be used, we may next inquire whether the closeness or remoteness of the connection, and hence the ease or difficulty of transition, between the idea of the temporal clause and that of the main clause in the same sen- tence finds expression in a similar way. The point about which the thought in a clause centers is usually the subject, the rest, predicate as well as modifiers, being subsidiary and auxiliary to it. Transition from one clause to the other will be easy when they have the same subject, and more difficult when they do not. And we find further confirmation of our views in the fact that identity of subject is much more common between the main clause and clauses introduced by postquam, ubi, ut, and simulatque,^ than between the main clause and the clauses introduced by cum. In the table which follows, all impersonal expressions have, of course, been omitted. Cum Sam© Different Subject Subject u 107 13 22 92 98 15 48 21 29 59 84 Percentage with Same Subject All Specific Conjunctions Same Subject Different Subject Percentage witli Same Subject Caesar Sallust Nepos Virgil {Aen. i-vi) Ovid (Jlfe^i-v) Livy (xxi, i)* 47- 37+ 48+ 24- 42 41 + 72 97 65 43 98 34 46 109 16 47 56 50 64+ 47+ 80+ 48- 64- 40+ *In Livy the comparatively small proportion of clauses with definite conjunctions that have the same subject as the main clause is in part due to his frequent use of impersonal expressions introduced by these conjunctions, especially by ubi. 1 Dum is the only specific conjunction whose clause more often has a different sub- ject than the main clause. This is largely due to the stereotyped character of many of the d«m-clause8, and the frequent use of such expressions as " dum haec geruntur," with which the main clause almost necessarily has a different subject, in the nature of the case. 270 J. J. SOHLICHEB Related to identity of subject as indicating ease of transition between the main and the subordinate clause is identity of tense and stage of action. A particular form of the verb carries with it a specific mental adjustment, which it has in common with the same form of other verbs. Figuratively speaking, we might say that a particular tense and mood of a verb fits into a particular mental mold, into which like forms of other verbs also fit more or less accurately, but all of them more accurately than different forms of those verbs. We see the operation of this principle in the leveling process which produces the sequence of tenses and the attraction of mood. Both of these phenomena owe their existence to the fact that like tenses and moods present a path of less resistance than would the use of different tenses and moods. Transition to the same tense, or to one belonging to the same time-sphere, or representing the same stage of action, is much easier than to a different tense or one representing a different time-sphere or stage of action. It is no doubt due to this fact that we find cum so much more frequently used to introduce a verb differing in tense from that of the main clause, than the other conjunctions, which we found associated with easy transition.' Additional light is thrown on the main subject of our discus- sion by a comparison between the cwm-clauses with the indicative and those with the subjunctive. The former are more closely akin to the clauses with specific conjunctions, which, like them, have the indicative. We find that these two kinds of clauses are included in a common preference by the same writers. Those who place their main reliance upon the specific conjunctions (Sallust, Virgil, Ovid) also have a relatively large number of indicatives with cum, while it is the writers making abundant use of cum, like Caesar and Nepos, who have also, relatively, the largest proportion of subjunctives. iHere, as elsewhere, an exception must be made of dum, "while," and the present indicative, the stereotyped character of which largely exempts it from the influences to which the other clauses are subject. The Tempoeal Cum-Clavs^ and its Rivals 271 Specific Cum Cum Conjunctions (total) (w. Ind.) 209 46 22 79 194 13 88 157 10 90 63 51 156 50 35 Cum (w. Subj.) Sallust Nepos Caesar Virgil {Aen. i-vi) Ovid (Met.i-v).. 24 181 147 12 15 In other words, the peculiar form of thought represented by the clause with cum, finds its extreme expression in the subjunctive with cum, the opposite extreme being represented by the clauses with postquatn, ubi, etc., with which cum and the indicative is much more closely allied than cum with the subjunctive. We reach the same conclusion when we consider the two kinds of clauses introduced by cum from a psychological standpoint. For most of the cwm-clauses in which the indicative is used, show characteristics similar to those which we have found associated with the clauses introduced by specific conjunctions. Take the cwm-clause of equivalent action [cum quiescunt, probant), for instance. We may in such a case put a sign of equality between the cwm-clause and the main clause. The two are cast in the same mental mold, as it were, and the mental effort expended on the one will serve for the other as well. The energies of the mind can thus be exerted in full upon the thought of the subordinate clause, since they are not distracted from it by that of the main clause, which coincides with it. The subordinate clause thus enjoys conditions which are favorable to its early and full maturity. The same is true of a clause whose action, while not equivalent to that of the main clause, yet coincides with it in time {quom pug- nabant maxume, ego turn fugiebam maxume). Like the clause of equivalent action, it has the indicative. Compare with these the temporal cwrw-clauses with the subjunctive, whose action in the case of the imperfect regularly extends beyond that of the main clause, while in the case of the pluperfect it lies entirely outside of it. What has been said of the cwm-clause of equivalent action, is also true, in a different way, of the postpositive cwm-clause (Vix eafatus erat, cum circumfusa repente \ scindit se nubes). This 272 J. J. SOHLIOHEK clause also gets much more of the speaker's attention than the ordinary cwm-clause of narration, for the reason that it contains the most important part of the thought, that which is formally the main clause being in reality of less importance and merely pre- paratory to it. But even when the cwm-clause merely adds a statement loosely to the main clause (what Hale calls a "forward- moving clause"), which is not of overshadowing importance, the fact that the main clause is already complete and can no longer distract the attention from the subordinate clause, gives the latter sufficient advantage to make its mood the indicative. In the case of the cwm-clause of repeated action, the very fact of the repetition has made the occurrence familiar to the speaker, so that it does not require the same time or effort to bring it to maturity in his mind, as would be required for a clause expressing a single occurrence. The repetition of the experience here, just as the greater concentration of the attention in postpositive cum, gives this clause an advantage similar to that enjoyed by the clauses with postquam, ubi, etc., as they have been described in an earlier connection. We may finally add to the list those clauses with cum and the indicative which are spoken of as "dating" or "expressing the exact time of" the action in the main clause. These definitions seem to mean something like the following. Two occurrences, A and B, take place, let us say, at the same time. A may be said to date B, however, only when A is, to begin with, definitely located in the speaker's and the hearer's mind, that is, when by extensive mental association with other known facts it has itself become accurately bounded and fixed. An example frequently met with in the grammars is this from Caesar B. G. vi. 12. 1: "Cum Caesar in Galliam venit, alterius factionis principes erant Aedui, alterius Sequani." Here the subordinate clause gives an event which was naturally prominent in the writer's mind and had, no doubt, often been used by him as a temporal landmark to measure from and refer other events to. Its action had conse- quently attained a clearness and definiteness which gave it a great advantage over other less prominent events, which passed through his mind less frequently. Hence the indicative. The Tempoeal Cum-Clavse and its Rivals 273 In external form the cwm-clauses with the indicative appear more closely related to the clauses with postquam, uhi, etc., also by the fact that more of them have the same tense as the main clause or, what for our present purpose amounts to the same thing, that the action expressed by them does not extend beyond that of the main clause. In a sentence like "Cum Caesar in Galliam venit, principes erant Aedui," the verb of the main clause tem- porally includes that of the subordinate clause, and the latter consequently does not, so far as the time goes, involve a distraction of the speaker's attention, which in resting on the action of the main clause also rests on that of the subordinate clause. In the cwm-clause with the subjunctive the relation of the tense to that of the main clause is decidedly the reverse of this. For there, with comparatively few exceptions, the action of the subordinate clause extends temporally beyond that of the main clause (imper- fect subjunctive), or lies entirely outside of it (pluperfect sub- junctive). Consequently, in the case of these, with his attention fixed mainly on the principal clause, as it generally is, the speaker's mind, in passing rapidly from one event to another, as it does in narration, is not able to bring the vagrant action of the subordi- nate clause sufficiently within his power to insure its prompt or full maturity in thought. And so he moves on, stepping from one main clause to the next, using the indicative, either with cum or with postquam, ubi, etc., in those subordinate clauses to which he can give the necessary attention in his progress, and leaving the others to be introduced by the vaguer cum, and con- cluded by the vaguer verb-form, the subjunctive, which stands for that mental half-light from which the thoughts expressed by them do not have a chance to emerge. VI The distinction in thought, however, between the two kinds of cttm-clauses, as, indeed, between the cwm-clauses as a whole and those introduced by postquam, ubi, etc., is by no means to be regarded as something absolute, or compulsory in its validity. For practical purposes, it is not even necessary that the distinction should appear in speech at all, as is shown by our own practice in 274 J. J. SOHLIOHEK English, where no visible difference exists between one kind of wAen-clause and another. And yet, while not a necessity of speech, the distinction is one of great stylistic importance. The avoidance by the poets of temporal cum, especially cum with the subjunctive, and their preference for the specific conjunctions, has already been referred to in the course of the discussion. It is clear that the latter clauses and cum with the indicative must have contained elements of meaning and the capacity for producing effects which served their particular purpose and harmonized with their habits of thought. These conjunctions were, as we saw, associated closely with fully matured, vigorously grasped, and clearly organized ideas, and their use served, no doubt, to stimu- late in the reader's mind that concrete picturing faith of a poem's contents, which makes such a large part of its successful reading. In poetry details are more important than in prose, and the sug- gestion, which the definite conjunctions give, of full attention to the subordinate parts of the sentence, has no small share in arous- ing the feeling that everything in the poem exists and is to be read and clearly realized for its own sake, and not merely as a means, which may presently be discarded, to some other end. But there are ways also in which the use of the definite con- junctions might serve a writer of prose. Sallust has been men- tioned as one who shows great preference for them. We can speak of only one effect produced by them which falls in with his peculiarities as a writer. There is no more prominent quality in Sallust than his apparent moral purpose. Moral earnestness in his day meant holding up the ideal of Cato the Censor and the good old times. And if we may trust the judgment of his critics, Sallust was, in language as well as in sentiment, a "fur Catonis." To the bluntness and avoidance of artifice, which he evidently cultivated, the temporal clauses with specific conjunctions might be of considerable service. Standing for full maturity and definite expression of every detail, they suggested a willingness on the part of the writer to give attention to all the facts, however inconvenient they might be, and to set them forth simply and honestly, and without manipulation. The maturity of thought associated with these clauses implies also that the writer is in no The Temporal Cuu-CijAVSE and its Eivals 275 hurry, but ready to stop long enough to allow the details to be clearly seen as they are. This is in strong contrast to the half- matured and slighted ideas of the subjunctive cwm-clauses, in which the mind of the writer seems always fixed on something farther on, which is of more importance to him. These latter clauses suit a style of rapid narration, like that of Caesar, just as the former suit the style of an author like Sallust, whose strength lies rather in the description of conditions. Tebke Haute, Ind.