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. ABSTRACT NOUNS IN HOMER By George Melville Bollino Professor Scott has returned, CP, XIV, 144 ff., to the question of the distribution of abstract nouns in the Homeric poems in a fashion which seems to me to add to the confusion already preva- lent. If that were all, the matter might be allowed to straighten itself out, as it would do in the course of time. But Professor Scott also charges a most distinguished Hellenist with what amounts to bungling dishonesty. His verdict runs: Maurice Croiset "de- serves severe criticism for taking over what he assumed was the correct figures of another and publishing them as his own." This is a serious charge brought against an eminent scholar, a man of blameless reputation, and before accepting Professor Scott's verdict it is right to consider the evidence upon which it is based. An examination of it will, I believe, free Croiset from any such impu- tation. Croiset published in his Histoire de la litterature grecque, 1 1, 384- 87, statistics of the use of abstracts in -crvvr), -it], -rvs for the pur- pose of showing their increase in the Odyssey as compared with the Iliad. The totals given were 58 for the Iliad and 81 for the Odyssey. In the Classical Review, XXIV (1910), 8-10, Professor Scott pointed out various 'errors in Croiset's statistics,' resulting in a change of the total figures to 79 for the Iliad and 81 (85) for the Odyssey. No comment was added, and indeed no comment seemed needed. This argument of "disintegrating criticism" had been tested, and it had proved merely that the scholar who had advanced it was an incompetent observer. The cleverest expres- sion of the Unitarian view of the situation was found by Mr. Shewan (CP, X [1915], 155): "It only required a count, as once at the English court it only required the weighing of the pail of water with the fish in it." 1 1 use the third edition of 1910, Professor Scott the second of 1896, and the passage goes back to the first edition of 1887. [Classical Philology XIV, October, 1919] 328 Abstract Nouns in Homeb 329 Such views I have never shared. To me the following argument seemed, and still seems, to hold good : First, the divergence between Croiset's statistics and the actual usage of the Iliad is too great to be explained 1 as due to inaccuracy; secondly — this merely for logical completeness — in the case of a man like Croiset there can be no question of intentional misrepresentation of the facts; and thirdly, there remains then no other possibility except the hypoth- esis that Croiset's figures present the usage of what I may call an Iliad proper; that is, a text from which parts regarded on other grounds as late have already been subtracted. At my suggestion one of my students, Mr. Erwin Esper, made a recount of the abstract nouns excluding the four "Odyssean" books (IKSFS2) and as an introduction to I, and by so doing reached results in sub- stantial agreement with those of Croiset. I was still puzzled by the fact that Croiset's meaning was not expressed with unmistak- able clearness. 2 The explanation of that fact offered below may not be entirely satisfactory, but that is the least part of the ques- tion. The interpretation of the statistics here suggested must either be the one intended by Croiset, or else we have an incredible series of coincidences. Pending the solution of this puzzle I did not encourage the publication of Mr. Esper's results, wishing also to review first the earlier treatments of the topic, especially that of Geppert. 3 In the meantime Professor Scott has himself compared Geppert and Croiset with results that are amazing. He writes (the italics are mine): "Croiset does not quote any authority for these figures, so that the assumption would be that he had made this investi- gation of the abstract for himself, but this is not the case, for I have found the source of the arguments and the identical figures in Geppert." 1 Not "excused," for I have never regarded either "carelessness in compilation or intentional deception" as an "excuse" for anything. 2 It is to be noted, however, that the obscurity turns on a single phrase. Had Croiset printed "qui appartiennent a Vlliade propre" instead of "qui appartiennent en propre al 'Iliade" his meaning would have been clear. 8 For that piece of bibliographic information I was indebted to the indefatigable energy of Mr. Shewan (loc. tit.): "The idea .... was as old as Geppert, and was constantly repeated, as by Bernhardy, Burnouf, Croiset, and Cauer." 330 Geobgb Melville Bolling To make the comparison for ourselves is not so easy a matter as would appear from Professor Scott's statement, for Geppert gives no figures at all. What Professor Scott means to assert is that computations based on lists of words given by Geppert will lead to 'figures identical' with those published by Croiset. That assertion is in absolute opposition to the facts. The computation may be made in various ways, for Geppert's lists are obscure. They are arranged (cf. p. 84) on the following principle: words peculiar to his Iliad, words peculiar to his Odyssey, words found in both these poems, then the words occurring in other portions of the vulgate. Some words in the last list have already been mentioned as occurring in his original poems; throughout plural and singular of the same word are handled as two words; for some words it is left an open question whether they should be included or not; with the suffix -rfc are mingled other words in -vs; one word is put down under those occurring in the original Iliad alone and after- ward listed as belonging to both the original poems; and finally single occurrences of some words have either been missed or inten- tionally passed over in silence on the ground that they stand in interpolations. Evidently the ways in which the computation can be made are many; but there is not one of them which will disclose a single coincidence between the figures for Geppert's original poems and the statistics of Croiset. Yet these are the figures which Professor Scott declares to be identical. To prove this I show in the accompanying tables and the appended notes all the possibilities, except that I have not counted any word twice, either because it occurs in both the "original" and the "late" parts of the poems, or because it is found in both the singular and the plural. 1 On the line marked Geppert I are given the figures for Geppert's original poems, which are according to Professor Scott the source of Croiset's statistics. On the line marked Geppert II are given for the sake of completeness (though they are in reality irrelevant) the figures for the vulgate editions of the poems. Here coincidence should be expected — supposing that each scholar had independently 1 Had the count been made in this fashion it would not have aided Professor Scott's case. Abstract Nouns in Homeb 331 ascertained correctly the number of occurrences in the vulgate. Such coincidence, even if perfect, would therefore contribute absolutely nothing toward proving that Croiset had 'taken over what he assumed was the correct figures of Geppert.' Fortunately even coincidence of this sort is almost entirely lacking. 1 I follow the order in Croiset which differs from that of Geppert. For -trbvt} the figures are as follows: Iliad Odyssey Both Croiset 6 14 6 Geppert I 12 6 2 Geppert II 13 15 5 Note. — Geppert mentions tytj/wcrvvT) both as peculiar to his Iliad and as found in both of his original poems. Correction of this will change his figures to 11-6-2, 12-15-5. Furthermore he ignores (perhaps intentionally, cf . under -tvs) the fact that tmroo-unj, /tavroo-wij, woft;/too-wi} are found once each (to 40, i 509, ir 233) in the Odyssey. Correction of this too yields: 8-6-5, 9-15-8. No two figures are identical. Yet it is from these that Professor Scott draws the only example of identity which he submits as evidence. His words are: "Croiset limits the number of abstracts in -abvi\ in the Iliad to twelve, which is precisely the number in Geppert's first list of abstracts in his Iliad." The number twelve for Croiset is evidently obtained by adding the figures of the first and third columns — a process as faulty in logic as it is impeccable in arithmetic. What Professor Scott must show is that 6+6 and 12+2 are identical figures. For -l-q the figures are as follows: Iliad Odyssey Both Croiset 21 32 17 Geppert I 23 13 8(11) Geppert II 31 28 11 (14) Note. — The variation according as the group dyyeAoj, ap/unit), Otowpowin is excluded or included. x The reasons will be understood readily: for the Odyssey it is the "personal equation"; for the Iliad it is due to the fact that Croiset does not include the whole vulgate. 332 George Melville Bolling Again no two figures are identical. Without comment I pass to the figures for -t6s: Iliad Odyssey Both Croiset 5 9 3 Geppert I 3 6 1 (4) Geppertll 5 10 2(5) Note. — The figures in parentheses include the other words in -vs. Geppert fists opx'? " Tv s under the Odyssey, regarding (cf. 1, 21) N 731 as demonstrably spurious. At last we find one pair of "identical figures," but even that one coincidence is not with Geppert I, as Professor Scott's charge alleges. To increase this by undoing Geppert's critical manipulation of dpxyvrvs would be simply to obliterate one indication of the fact that Croiset's work is not dependent on Geppert. Even that, as already pointed out, would not strengthen Professor Scott's case. That demands a coincidence between Croiset and Geppert I, for which there is neither here nor elsewhere a single shred of evidence. Professor Scott's charge meets with another difficulty. It is perfectly obvious that anyone using Geppert must see " that ' Iliad ' in Geppert did not mean the poem of that name but meant only portions" of that poem. To avoid that difficulty Professor Scott sets up the hypothesis "that Croiset obtained his figures not from Geppert but from some intermediate source in which [this] fact was overlooked." This ad hoc assumption, however, need not be dis- cussed, because the fact it is advanced to explain (the alleged identity of Geppert's and Croiset's figures) has been proved not to exist. Until it can be shown that the pairs of figures 6-12, 14-6, 6-2; 21-23, 32-13, 17-8(11); 5-3, 9-6, 3-1(4) are identical it must be held that Croiset's figures were not obtained, either directly or indirectly, from Geppert. The idea that there was a progressive increase in the number of abstract words, and that contemporary literature would reflect this development, was of course not original with Croiset. To judge from Geppert (p. 83) it was a commonplace in 1840, when he used it to demonstrate the difference between "Homeric" and "post- Homeric." No doubt it figured also in Pazschke's De nominibus abstractis apud Homerum, which according to Terret appeared at Abstract Nouns in Homer 333 Halle in the same year. I have not the means to follow the trans- mission of that idea in all its ramifications and should consider it futile to do so. Mr. Shewan (loc. cit.) has singled out two prominent exponents of the idea. Both are among the sources of Croiset (Preface, pp. xxxixf.), and their treatment of the question may be expected to give a fair picture of the form in which the idea reached him. Bernhardy, Grundr. d. gr. Litt., II, i 3 (1877), p. 157, makes a general reference to Geppert for detailed work on linguistic differ- ences within the Homeric poems but, when he comes to speak of the abstracts, cites neither Geppert nor any other authority. What he says (p. 179) is as follows: Im Wortgebrauch wie in Phrasen und Formeln neigt der Stil [der Odyssee] nach Art der geistesverwandten hinteren Buecher der Mas, zur Abstraction und Kuenstlichen Metaphern, waehrend er von der naiven Sinnlichkeit sich entfernt; nicht minder wechselt er aber auch nach Buechern. Da bei Homer die Zahl der fuer abstrakten Wortsinn bestimmten Endungen klein ist, so darf man nicht uebersehen dass solche weniger in den aelteren Buechern der Mas, haeufiger in den hinteren Gesaengen der Odyssee vorkommen: besonderen auf -tvs (ptxrraicTvs) und -avvv\ kXiwtoctvvt), rapfioo-vvri, tcktoctv- vdmv. Das unpassendste Wort der Art ist in *> 167 wo\.vKepSciyiXopvpir) (cf. vorirj), Xo<£ii7 (cf. Ludwich at r 446 for ancient disputes about the accent), and virepreplr]. 1 The matters mentioned in the notes appended to the tables are such as might naturally be ignored in a summary which was expressly stated to be incomplete. Abstract Notjns in Homer 335 Abstract nouns in -rus are found as follows: Iliad Odyssey Both Croiset 5 9 3 Esper 4 9 3 Note. — There is found besides ixavrurrk * 622. Here for the first time the hypothesis seems at fault, since Croiset's figures are correct for the whole Iliad and not for the Iliad proper. The fact must be weighed against the satisfactory explanation which the hypothesis provides for the figures of -vyri, qui est toujours remplace" dans Ylliade par 336 George Melville Bolling 0yfa ou 6@os (0v£is ne figure que dans la Dolonie)." Finally it is very noteworthy that Croiset does not here take into consideration the fact that the Iliad is longer than the Odyssey. He treats the poems as if they were of equal length, which is approximately true if Ylliade is taken in the sense which I have ascribed to it. There remains the question of how far Croiset's presentation of the subject can be taxed with a lack of clearness and of how such a lack of clearness can be explained. That question turns on what one has in mind when one speaks of the Iliad. A rather curious result of the Neo-Unitarian movement has been the apotheosis of the vulgate as established by Wolf. It is now apparently main- tained that the Iliad as written by Homer is the 15,693 lines that Wolf printed — no more and no less. But in the eighties of the last century that usage was not established. The word Iliad did not then have to be taken thus au pied de la lettre, but could be used to designate the genuine poem, disregarding its later accretions. Even then it was better to be precise and indicate what portions of the vulgate were thus discarded. This Croiset should have done; and no doubt he would have done it, had it not been for his intention of publishing elsewhere a detailed treatment of the subject. The alternative to believing this is to maintain that a blundering count led by accident to the results that might have been attained by setting aside the five books specified. There is, however, one blunder which Croiset did make, and that one Scott did not detect. It occurs in a footnote of the first edition, and was corrected by Croiset himself in the second edition. The total of Croiset's figures for -abvrj, -ir/, -rvs is actually: Iliad 58, Odyssey 81; and these totals are so given in the second and third editions of his work. Cauer (Burs. Jahresbb. (1902), p. 119, and Grundfragen, p. 393) quoted 39 and 81 as the totals. Scott noticed the discrepancy, apparently regarded it as a blunder of Cauer's in copying or printing, corrected it in his article in the Classical Review, and still refers to it as 'the final form of the argument.' As a matter of fact the figure 39 was the one given by Croiset in his first edition, and it was not a misprint; for comment is made upon the fact that the total (39) for the Iliad is less than half of the total (81) for the Odyssey. Abstract Nouns in Homer 337 The explanation of this blunder is simple and clinches my argu- ment. The fact that the addition is given in a footnote shows that it was an afterthought. In a moment of forgetfulness Croiset fancied that 58 was the total for the whole Iliad and subtracted from it the nineteen words (listed above) which occur in GIKSFQ but not in the other books of the Iliad. I have intentionally avoided any discussion of the merits of the argument itself, and have simply tried — it is the thing that should be done first in all polemics — to understand it as its author meant it. One may draw, as far as the present article is concerned, what conclusions he will from the distribution of these abstract suffixes in the Homeric poems. But it will be, I trust, impossible to argue hereafter either that Croiset was incompetent to compile such statistics, or that he was "taking over what he assumed was the correct figures of another and publishing them as his own. " The Ohio State University