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 by JSTOR. 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.jstor.org/participate-jstor/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. Book Reviews 367 The remarks on phonology (pp. 6-11) fail to cover several phenomena of interest that appear in the texts and in examples given elsewhere in the book. The statements made in paragraphs 9, 13, 14, 15, 17, 19, and 20 should have been accompanied by illustrations. The absence of examples here is in strong contrast to their copiousness in other parts of the grammar. But the author has purposely laid little stress upon rules of phonology and has entirely abstained from the alluring occupation of comparing Bontok with other Philippine languages. As an intensive first-hand study and exposition of the phenomena of a single Philippine idiom, Dr. Seidenadel's work stands unique, and sets a standard of thoroughness and accuracy for which all scholars will be grateful and which may well be followed in future monographs on individual Philip- pine languages. C. E. Conant The University op Chicago March 4, 1911 Syntax of Early Latin. By C. E. Bennett. Vol. I, The Verb. New York: Allyn & Bacon; Leipzig: Theodor Stauffer, 1910. Pp. 506. $4.00. Professor Bennett's book deals with a field which has long been crying for cultivation. Its predecessor, Holtze's Syntaxis Priscorum Scriptorum Latinorum, is now fifty years old (published in 1861-62). Professor Bennett is to be congratulated on having accomplished a heavy and important task. The mass of material covered is large. A great many opinions are passed in rapid review. The discussions are on the average as full as could be hoped for in a general volume of such compass. The bibliography given is considerable, and well chosen. The style is lucid, flowing, and interesting, as always in Bennett's work. The points of view are a great improvement over Holtze's — to whom, however, now that we are parting with him, it ought to be said that many a man has had reason to be grateful. The store of examples is extensive. The book deserves wide reading. The publishers, with their usual excellent taste, have done their part to make this easy. Having thus summed up my general opinion, I must hold myself free to express assent or dissent in detail. This is facilitated for me by the fact that Bennett in a large number of places cites my monographs and the syntax of the Hale-Buck Grammar, with assent often, to be sure, but also with fre- quent dissent, and the discussion of his reasons. His work and mine have, in fact, lain largely in the same field. We are thus natural critics, the one of the other. The general plan is indicated by the chapter headings, which are as fol- lows: I, "Agreement, Voices, Impersonal Verbs, Omission of Verbs"; 368 Book Reviews II, "Tenses of the Indicative (Especially in Principal Clauses) "; III, "The Indicative in Subordinate Clauses"; IV, "The Subjunctive in Principal Clauses"; V, "The Subjunctive in Subordinate Clauses"; VI, "The Imperative"; VII, "The Infinitive"; VIII, "Participles, Gerund, and Su- pine, etc."; IX, "The Sentence-Question." Already, the headings betray a lack of architectonic care. IV and V correspond. To these, again, III corresponds. But II does not. Further, it turns out that II deals not "especially," but only with the indicative in principal clauses, and that the important subject of the behavior of the tenses in subordinate clauses is not treated anywhere in the book. The same lack shows itself in the treatment of the subjunctive. There is no general statement of the tense-forces in the independent uses. There is no discussion of the rationale of the phenomena known as the "sequence of tenses," such as I gave in my "Sequence of Tenses," A.J. Ph. IX, No. 2 (1888). For this important and much-discussed field, too, a single reference only is given, namely to Wetzel, De consecviione temporum Plautina el Terentiana, 1888. To my mind, a general discussion of tense should precede the detailed discussion of the moods. There are certain powers which are common to many parts of the verb. For example, there is clearly a use of the perfect to express energy, promptness, decision, and the like, through the repre- sentation of the act as a finality. I agree with Bennett that Elmer has not made out his point against Madvig; but it seems to me nevertheless sure that the perfect subjunctive in prohibitions originally expressed energy, etc. (Bennett seems to take it as merely aoristic.) But the same energetic power is possessed by the perfect subjunctive in statements, the perfect indicative, the future perfect indicative, the perfect passive participle, and the perfect passive infinitive. Thus perii, Trin. 1081 (p. 45), periisse, Most. 349, perieris (a wish), Men. 295 (p. 195). The use of the perfect participle and infinitive in this sense (thus monitos, Cic. Cat. ii. 12, 27, and interfectum esse, Cic. Cat. i. 2, 4) is very common in all periods. These identical phenom- ena should be brought together, as I have brought them in half a page in the Hale-Buck Grammar, §490. Bennett speaks of the tense nowhere except for perii, where the explanation that it is like memini and odi seems to me to miss the very point for which the tense was used. I do not in the least agree with Bennett's treatment of eras, erat (p. 34) as aoristic. Of course there is, quite as he says, no "continued action" in erat; but neither is there in any other tense of sum. Sum is not a verb of action. But erat expresses situation, as the imperfect of any verb regularly does. There is no better touchstone of tense-feeling than in these same erat sentences and clauses. Your Italian washerwoman of today still has it, using era and fu just as the Romans used erat and fuit. — Neither do I in the least agree with Bennett's view (which is the general one) that the aoristic perfect of verbs in general is "momentary" (pp. 10-17). It simply presents Book Reviews 369 an act in summary, in the lump, without any question of "duration" or "momentariness" (neither of which terms has any proper place in grammar). The conception of parataxis is employed, without explanation, on p. 60, at the opening of the treatment of the indicative. On pp. 208 and 209, under the heading "subjunctive in subordinate clauses," it is fully discussed. Under the imperative, where it is important to notice it, it is not mentioned. The general statement should be given where the conception is first employed. The treatment of the subordinate indicative clauses begins with one of the most complex of all, the "conditional sentence." There is no treatment at all of the indicative gm-clauses, except incidentally under the subjunc- tive (pp. 137, 142) in the use which I have noted and named (Cum- Constructions, pp. 104 and 113-119), the "tacit causal or adversative qui- clause." These are the only indicative constructions referred to in the index under the heading "relative clauses." Yet a clear comprehension of the indicative relative clauses is a great help in the comprehension of many of the indicative clauses with conjunctions. I miss, too, a treatment of the generalizing indicative clause (H.-B. Gr., §579, n. 1), the lack of an understanding of which common construction causes many a writer to go astray, as Lindsay, who (note on Capt. 462) is surprised at a construction which is not only regular but constantly occur- ring (indicative gm-clause, in spite of a causal implication, because the clause is generalizing, and so, in effect, an assumption, not a predication). So much must suffice as regards the general structure of the book. We come now to some of the controverted points at which I agree with Bennett. I have already spoken with approval of his rejection of Elmer's treatment of the prohibitive. The same holds for his retention of the cate- gory of the potential of possibility (I have thus to employ tautology because Bennett follows the almost universal but entirely unscientific and altogether inconvenient practice of naming two quite distinct forces of the subjunctive by the one term potential 1 ) against Elmer's attack; though I do not, with Bennett, limit the use to examples like videas. I agree with his opinion that a true indirect indicative question exists in early Latin (as also, here and there, in the later poets). I take pleasure in his demonstration that Morris is wrong in explaining the first subjunctive in combinations like velim eas as due to the attraction of the second; though he might well have added that Greek has the exactly corresponding /3ov\oC/j.rjv av (with infini- tive always), for which Morris' explanation is of course impossible. 2 I i The necessity of having two potentials, a "can-could potential" and a "should- would potential" (Bennett, passim) is a practical confession that "potential" will not serve, and that some such term as my "ideal certainty" is necessary for the "should- would" idea. And, in any case, imagine a naturalist using the same name for types as far apart as these ideas generally are! ' Morris, Principles and Methods in Syntax, discusses with great fullness the proper methods of procedure. In the entire book, he offers an explanation of his own in only one case — the one referred to — where he seems to be quite wrong. Nothing better illustrates the weakness of Morris' opposition to the comparative 370 Book Reviews am glad that Bennett agrees with me as against Elmer that the subjunctive of the first person with opinor is volitive. I of course heartily agree with his direct acceptance (p. 303) of my doctrine with regard to the cwm-clauses, and (p. 338) of my explanation of the "iterative" subjunctive as descended from cwwi-clauses with the subjunctive; though I differ with regard to some of the examples cited. I am pleased that he accepts and employs my term ' ' descriptive clause ' ' as better than ' ' clause of characteristic' ' ; though he does not point out the strongest reason for it, namely that the clause in question frequently states, not something permanent, as a clause of "characteristic" should, but a merely temporary condition. I am glad that he also, for the commonest type (that which corresponds to an independent statement of fact) accepts and uses my term "descriptive clause of fact." I am glad that he uses my term "determinative clause," just as I was sorry to see that in the -last edition of his Appendix (under the title of The Latin Language) he omitted the chapter which Elmer had supplied, from my teaching at Cornell in the days when Elmer was my younger assistant, upon indicative relative clauses — including the "determining" clause. 1 I was also glad to see that Bennett, in treating the descriptive clause, recognized that my C urn-Constructions dealt almost as fully with relative clauses as with cum- clauses, and accordingly gave my book as his principal reference ("see especially," etc.). And I am particularly glad to see that he recognizes my tenet that a number of the subjunctive constructions are of composite origin — a tenet which has not appeared in any of his writings hitherto. I also agree that his distinction of the "stipulative" subjunctive (p. 263, from Trans. A.P.A. XXXI, 223 ff.) as a development of the volitive subjunctive (as in Aul. 458, lege agito mecum, molestus ne sis, "on condition that you only let me alone," p. 265, bottom) is sound and helpful. I am most heartily in accord, too, with Bennett, in that he treats syntax in this book, as he always has treated it, as a matter of psychology, and not of metaphysical categories. 2 He cites with approval (p. 147) my "A Century of Metaphysical Syntax," Congress of Arts and Sciences, St. Louis Universal Exposition, 1904, Vol. Ill of the Proceedings, to which might be added my method than the fact that, in this case, he would have to explain a Greek (and an English) construction by the very way which he rejects for an apparently exactly cor- responding clause in Latin. 1 On p. 288, Bennett rightly expresses my conception of the determinative clause : "Hale .... defines the determinative clause as one telling what person or thing is meant." But the phrase just preceding, " 'determinative' clauses, which simply add an- other fact or item with regard to a person or thing," is not accurate. Thus the qui- clause in Cicero, qui . . . . , adds something, but is not determinative. A proper name is se(/-determined. 2 Yet his category of "contingency" as explaining the "can-could potential" and the "should-would potential" (p. 205) is a metaphysical one, established in our grammars by the application, in the latter part of the eighteenth century, of Wolff's categories, in the Ontology, to the moods of the verb. Book Reviews 371 ' 'The Heritage of Unreason in Syntactical Method," address before the English Classical Association, 1907, Vol. V. It was not previously understood how deeply conceptions derived ultimately from Hermann's application (1801) of Kant's modal categories to the moods of the Greek verb are still imbedded today in the grammatical consciousness of the great majority of workers in mood-syntax. Bennett's predecessor Holtze is full of this metaphysical syntax. Thus (II, 142) he explains non .... redderem? Trin., 133, as rem tanquam cogitatam. Bennett, like others except myself, has hitherto always treated all sub- junctive questions as deliberative, while I have distinguished deliberative questions, questions of obligation or propriety, potential questions, etc. He now, without mention of my procedure, condemns the idea that all subjunctive questions are deliberative, and has "questions of duty or fitness." I have pointed out (H.-B. Gr. 502, 1) the existence of a volitive descrip- tive clause. Bennett, without mention of this, now has a jussive descriptive clause (p. 294), as well as (quite rightly) an "optative descriptive clause" (p. 295), which also I have included in syllabuses for class use. For the former, of which he says the material is scanty, he cites M.G. 140, Cat. Agr. 21. 1 and 2, and, "less certain," Bacch. 67. I should put only the two ex- amples from Cato here. The last is possibly volitive, but I should incline to take it as an anticipatory descriptive clause. The first I should call a "clause of plan " (our common phrase " clause of purpose " gets us into trouble by its narrowness). The general introduction upon the mood-forces of the subjunctive (pp. 145-61) is good; though I was sorry not to see a mention of Delbriick's explanation (N. Jahrbiicher, 1902, pp. 33 ff.) of the rise of the "potential" power out of the optative, or of my view, of which I shall speak presently. The summary of Stahl's general reasoning is well included. One of the leading forces in my mood-system is that of obligation, propri- ety, or reasonableness. I find that the subjunctive has it in Latin, and that the optative has it, in large use, in Greek, with the subjunctive in much smaller use. I believe the Latin use to be mainly of optative origin, but partly of volitive. The category goes much beyond Delbriick's "prescrip- tive," not being confined to general expressions, and appearing often with av. I find abundant examples of the latter type in Greek, as well as with the bare optative. A force of this kind has been recognized in the past tenses in Latin, being explained as a "jussive" thrown back into the past. I find that it is used freely in the present also. The category was first laid down by myself, but is commonly spoken of as Elmer's (who gave it a different origin). As he has explained in print, he heard it spoken of so frequently in my class- room that he supposed it to be a commonly recognized category. Bennett has never had this category in his writings. Nor has he now in the new book. But he has the same thing under another name, "the sub- junctive of duty or fitness." Moreover, in his translations of examples 372 Book Reviews under it (p. 180), he says, "in the third person" (why this person only?) 1 "the meaning often is .... is it natural that he should do" or "be," .... "how is he likely to know me?" .... "what is it 'natural' or likely to be except this?" This is nothing in the world but another of the leading forces of my system, namely the subjunctive of natural likeli- hood — with the exact adjectives corresponding to the two parts of the name which I have given to the construction, and the exact derivation. Bennett's book has completed the process of making me orthodox, without any change of opinion on my part! The importance of the two forces lies partly in the fact that they make the bridge between the optative force and the "potential" force — wish to obligation (or possibly the opposite), obligation to natural likelihood (com- pare English "should" and "ought," which are used to express both ideas), natural likelihood to possibility in the one direction, and to ideal certainty in the other. This I have shown in the paper entitled "Leading Mood-Forces of the Subjunctive and Optative in the Parent Speech," P.A.P.A. XXXII, 1901, and in the Hale-Buck Grammar, 459. Bennett has not mentioned either of these, or my categories or views on the subject. He regards the declarative form in the imperfect or past perfect as a "jussive" thrown back into the past. The interrogative uses in the present (he sees no others) he traces to the deliberative, which of course we agree is of volitive origin. The corresponding question in the past he makes to be an outgrowth of this. The interesting result is that in Trin. 133 the question and answer have different names, the question non redderem being one of "duty or fitness," while the answer non redderes is one of "past obligation." I also assign to the subjunctive of obligation or propriety clauses like those in nihil est iam quod tu mihi suscenseas, Merc. 317, non iusta causa est quor curratur, Poen. 533, non iusta causa est ut vos servem, Capt. 257, ego te meru- isse ut pereas scio, Poen. 1407, and the dignus qui construction, as in est digna res ubi nemos intendas, Eun. 312 (originally from a type like "it is a worthy thing; you should put out your best efforts upon it"). Bennett calls the first example interrogative (p. 241), and the second a substantive clause developed from the deliberative (p. 248), puts the third under "miscel- laneous" substantive clauses, under the more general heading "expressions developed from the volitive" (p. 242), calls the fourth a substantive clause developed from the volitive (p. 240), and calls the last a clause of purpose (p. 260). My treatment appears to me both truer and simpler. Some of the principal points where we differ more radically are as follows: Bennett, in taking up the subjunctive, gives a number of examples, with- out citation, from Sanskrit, Greek, Germanic, and Slavic. This looks as if he were really going to employ the comparative method for possible light upon *SE: Ecquid sentis? MED: Quidni sentiam? Non potest haec res ellebori iungere optinerier (Af en. 912) ; CV: Nostin ? L Y : Quidni noverim ? (Cure. 423) . Similarly with ubi (Amph. 748). Book Reviews 373 Latin. But nothing of the sort is done. Only one subordinate example from any of these languages is mentioned by him (p. 271, from Delbriick Vergl. Synt. II, p. 401). The results which Bennett reaches at many points are in direct opposition to the facts in the languages that possess two distinct moods, as against the Latin mixed mood called the subjunctive. I will point out some of these contradictions, confining myself to Greek. It should be borne in mind that volition is expressed in Greek by the sub- junctive, without av or *«; "futurity" (anticipation, as I have called it) by the subjunctive with or without av or «e; wish (when without any intro- ductory word) by the optative without av or /«; and the potential idea (which in Bennett means the "can-could" idea or the "should- would" idea), by the optative with or without av or zee. 1 In what I say below, the word "optative," unless otherwise modified, will mean "expressing a wish." Bennett makes the clause of purpose of volitive and optative origin, adding that "the share of the optative in the development of the purpose clause is also probably as great as that of the volitive" (but the optative is rare in this clause in Greek 2 ); the subjunctive after antequam and priusquam the result of attraction after an imperative, subjunctive, or infinitive (the corresponding construction in Greek is the subjunctive, with av or *c, which is thus anticipatory; and attraction into the subjunctive is not a recognized or probable usage) ; the sub j uncti ve with dum, ' ' until, ' ' an optative (the corre- sponding construction in Greek is the subjunctive with av or ice, which is thus anticipatory) ; the subjunctive of indirect discourse an extension from the deliberative question (Bennett rightly makes the deliberative question to be a volitive subjunctive, while the mood of indirectness in Greek is the optative) ; the subj unctive of the indefinite second person singular the result of attraction in combinations like quos quom ferias, tibi plus noceas, Pseud. 137, the second verb being a potential (the construction, as far as it goes, looks like the generalizing construction in Greek, for which the mood is the subjunctive, not the "potential"). The clause of fear he makes to be optative. Against my view of its volitive origin (thus originally, "I don't want you to have an accident; I am afraid," which comes to mean "I am afraid that you may have an accident"), he says, "this view seems psychologically less natural than the one above indicated. A state of fear is quite compatible with a wish, but hardly with a peremptory order." But he himself tells us, on p. 162, that "often the force [of the volitive] is so weakened or modified as to convey simply a warning, admonition, or request." At any rate, what would he do with Greek, in which the clause of fear is regularly, not in the optative, but in the subjunctive, without a modal particle, and thus surely volitive in 1 Bennett speaks of the particles under the Greek subjunctive, but not under the optative. — My statement above is meant to conform to the freedom of Homeric Greek. 1 There are but three cases in Homer attested by the MSS. I accept them. Stahl, with most Homeric workers, emends the first two, and explains the third as indirect. But the number is in any case very small — three against hundreds. 374 Book Reviews origin? He would have to show us why that which was psychologically unnatural in Italy was psychologically natural in Greece. Clauses like those in nihil est quo me recipiam, Capt. 103, and est unde haec fiant, Ad. 122, Bennett calls relative clauses of purpose (p. 259), and argues against my view that they are mainly of potential origin. But clear po- tential clauses of this type are frequent in Greek. Thus Imroi 8' oi Trapma-i Kal ap/juira, tS>v k' liri/3airjv, II. 5, 192; SXX' ovSk fj.-qv vavs 2evyovres, Eurip. Hel. 1047; a. /Av ovv aAAovs «X eT£ oltrrura/ av SoCryre avrovs, Xen. Cyr. 4, 5, 47. I should not agree with Bennett's readings at every point. Thus he reads ostentat in Enn. Ann. 262 (414 Muller) . The indicative does not make easy sense. The weight of MS authority is heavily for ostendat, which (The Anticipatory Subjunctive in Greek and Latin, University of Chicago Press, 1894, p. 66) I take as anticipatory, "this is the day when the greatest glory shall reveal itself to us, whether we live or whether we die." Cf. "now the day was come when the queen should bear a child," Malory Morte d' Arthur 1, iii. But in Bennett's system there is no anticipatory subjunctive. A number of examples seem to me to be placed under wrong categories. Thus (p. 164) velitis iubeatis in the formula for the adrogatio in Gell. v. 19. 9 I should take as interrogative, corresponding to the ordinary velim, "I should like," velis, "you would like." Note the ita vos rogo at the end. Vel Mud perdas vel amicum amiseris, Trin. 1054, belongs, not on p. 166 ("jussive"), but on page 200, for which it is a perfect example, illustrating my theory ("An Unrecognized Construction of the Latin Subjunctive," Class. Phil. I, 21 ff.) that there is a generalizing second person singular subjunctive of statement, corresponding to an indicative second singular (perdas = perdis, except that the mood of the former calls attention to the idea of generality). It is balanced against fit in 1051, and exoritur in 1053 (I miss also in the list on p. 200 the example invenias of 1052, in the same context in Plautus, as also bibas, etc., Ter. Ph. 342) . Bennett, let me say in passing, really accepts the new category, translating, p. 200, by "you never do hear," "you scarcely ever find," while appearing, p. 201, to give a different solution for it. But he mistakes mine, which includes his, speaking of it as "formal assimilation." The explanation which I gave, op. cit., pp. 39 ff., was twofold. First, through use in the generalizing second singular indefinite the mood had come to seem the sign of universality, and so seemed appropriate here. Secondly, it often occurs, in the same second singular indefinite, in subjunctive questions of perplexity, in questions of obligation, propriety, or reasonableness, and in statements of natural likelihood, possibility, or ideal certainty (Bennett's "potential"). My explanation then includes his, namely that the construc- tion is "simply an attenuated ' should-would ' subjunctive." Bennett says, Preface, p. Ill, "figures in parentheses indicate the num- ber of occurrences of the usage which they accompany. Examples without figures, unless accompanied by 'e.g.,' are intended to represent the com- Book Reviews 375 plete material under the given topic." If so, there are errors. Thus three examples of the future indicative with antequam or priusquam are given from Cat. Agr. on p. 105, without an "e.g." But there are 4 (add 161, 2). Again, 7 examples of the subjunctive with the same connective are given from the same work. But there are at any rate 10 (add 143, 2, a second example from 53, and again a second from 113, 1). Similarly, p. 305, Bennett gives 6 examples of dum with the subjunctive from Cat. Agr. But there are 7 (add 103). On p. 97 he gives 15 for the number of examples of the future indicative with dum "so long as." But his own citations count up 16. They include Cat. Agr. 30, usque dum habebis, praebeto. To this should be added usque dum habebis, dato, ibid. (Further, dum poterit, 90, given on p. 97, under "as long as," seems to belong under "until," p. 98.) It is im- possible to collect and handle so large a mass of material without mistakes. But omissions so considerable as these somewhat shake one's confidence. Yet they may all be due to some accident within the limits of a single general group of constructions. The remaining chapters must be left without notice. The book is very free from misprints. I have noted only inter cedit for intercedet (p. 98, near bottom) and argentun for argentum (p. 35, near top) . William Gardner Hale Index Verborum Vergilianus. By Monroe Nichols Wetmore. New Haven: Yale University Press; London: Henry Frowde, 1911. Pp. x+554. $4.00. In 1904, Professor Wetmore published a Doctor's dissertation entitled The Plan and Scope of a Vergil Lexicon, a work which was widely dis- tributed and favorably reviewed in various periodicals. In 1909, H. Merguet, the well-known compiler of lexicons for Cicero and Caesar, who had been sent a copy of the above dissertation in 1905, proceeded to publish a Lexikon zu Vergilius. It is a pity that the elder scholar had not indicated his intentions, for he would either have spared Professor Wetmore the trouble of preparing some thousand pages for the press or, by collaboration with him, would have produced something far better than the present Lexikon zu Vergilius. No law compels courtesy on the part of competing scholars; still, there is no law against it. Kalinka, in a not altogether favorable review of Merguet (Berl. Phil. Woch. [1910] 1038) expresses the wish that Merguet and Wetmore would join forces and give us a perfect lexicon of Virgil. But the time for such a wish was the past. Despite Merguet's publication, Wetmore's Index fills an important gap. The old index of Erythraeus (1583), appearing after various revisions in Vol. VI of Heyne's edition, can no longer, after a century of work on the text of Virgil, hold the eel of science by the tail. Besides, it is extremely defective, especially in citing the minor words, whereas Wetmore registers all