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. THE SIGN OP INTERROGATION IN GREEK MINUSCULE MANUSCRIPTS By Charles Brewster Randolph The punctuation in Greek minuscule manuscripts strikes one as varium et mutabile semper, if not within single manuscripts, at least when several are compared, and this is doubtless one of the reasons why marks of punctuation are so, seldom recorded in the critical editions of Greek authors. Sometimes a sign takes on more than one function in the same manuscript, as, for example, the comma and the point in its low and its middle position in the scholia of Aristophanes in the Ravennas manuscript. 1 Frequently, on the other hand, it happens that in spite of the very great number of punctuation-marks in a manuscript, in many instances where punc- tuation is required it is altogether missing. 2 It is perhaps impossible to arrive at any broad and definite conclusions regarding the use of the high, middle, and low point, and even the comma, but it is possible to detect in many manuscripts a definite rule regarding the use of the sign of interrogation — a rule whose employment, if we may judge from its frequency in the limited number of manuscripts .of which facsimiles are at our disposal in America, appears to have been very common, if not the general practice in the writing of minuscule manuscripts. The paleographers, like Wattenbach and Gardthausen, have little to say about punctuation by points. Regarding the use of the question-mark they inform us that a sign of interrogation first occurs in the eighth or ninth century; that it has several forms; that the point and comma in combination (our semicolon), which is the commonest sign of interrogation, occurs also in places where it does not denote a question. 3 Zacher, in the preface to his edition of the 1 Rutherford Scholia Aristophanica, I, xxix. 2 See, for example, Tiachendorf Monumenta Sacra Inedita, Nova Collectio, II, xxxxi. 3 Wattenbach Anleitung zur griechischen Palaeographie (1895), 60, 121 ; Gardthau- sen Griechische Palaeographie, 276. [Classical Philology V, July, 1910] 309 310 Ohables Beewstek Randolph Equites of Aristophanes, 1 observes that all the Aristophanes manu- scripts employ the point and the sign of interrogation, but often omit both, and often put the point after a question instead of the sign ; . In examining a manuscript of Aristophanes some time ago I was struck with the persistence, as it were, with which the question- mark, disappearing after indicating one or two questions, and apparently given up by the scribe, appeared again, then apparently dropped out of use, and reappeared. I began to observe the form of the question and compare this with the sign at the end. A score of tests at the beginning of the Aves gave a good indication that there was method in the use of the sign. The principle that I want to establish first is this: The sign of interrogation does not follow questions introduced by an interrogative pronoun or adverb (for example, rln, -rrov, wm) . These I shall speak of as "word-questions." I first arrived at this conclusion by examining the punctuation of all the questions in a considerable portion of four of our best minus- cule manuscripts. These were Ravennas (R) and Marcianus 474 (V) of Aristophanes, Laurentianus xxxii. 9 (M of Aeschylus and L of Sophocles), and Oxoniensis Clarkianus 39 of Plato; to which I added a Harvard manuscript of the gospels, thinking that in the Bible if anywhere the scribe would understand what he was writing, and the manuscript would therefore exhibit faithfully his system of punctuation. This examination yielded results as follows: Manuscript Symbol and Author Parte Examined Word- Questions Times ; occurs Ravennas (XI)* Marcianus474(XII) Laurentianus ~) xxxii. 9 (X or [ XI) ) Ox. Clark. 39 (IX).. Harv. Gospels (XII or XIII) R of Aristophanes Vof Aristophanes M of Aeschylus L of Sophocles B of Plato Aves entire Aves entire Prom. Vinet., Agam. Antig., O. C. 1-525 Euthyphro, Lysis About 80 MS pages 197 199 74 118 100 46 7 2 2 Totals 734 11 * The Roman numerals indicate the century to which the manuscript is assigned. 1 P. xix : Interpungwnt omnes (codices ) puncto • et signo interrogationis ; sed utrum- que saepe omittunt et in interrogatione punctum ponunt. The reference is to Zacher'e revision of Velsen's edition, Leipzig, 1897. Sign of Intebbogation in Gtbeek Manusobipts 311 It should be remarked that in these and all the other manuscripts that I shall cite in support of this principle the question-mark is in use, and has no function aside from that of indicating a question. Facsimiles of the first four manuscripts are accessible in photo- typie reproduction, and for the fifth I had the original manuscript. The only other complete manuscript facsimile accessible to me in which the point and comma appear commonly in combination was Parisinus 2934, 2 of Demosthenes, in which the sign plainly has not interrogative value. 1 I therefore turned to the examination of such facsimile sheets as I could find containing passages long enough to provide a test. Fortunately I was able to avail myself of photographic facsimile sheets in the Harvard Classical Library containing passages of con- siderable length from twenty minuscule manuscripts, to which I have added seven photographic facsimiles of the Aves of Aristophanes, which Professor John Williams White kindly loaned me. For con- venience I have classified these by centuries. 2 The names of these twenty-seven manuscripts, with the results of my examination of them, appear in the table on pp. 312 f. 'See the full discussion of this below, pp. 316 ff. 2 The classification by centuries has no particular significance. It is a difficult or impossible task, especially in a photographic or phototypic facsimile, to determine how much of the punctuation is by the original hand ; this matter must be carefully worked out in the case of each manuscript, and this paper does not take into account the hand who wrote the sign of interrogation, attempting to show simply what we may expect to find in a manuscript today. When I speak of "the scribe" in connection with the punctuation I mean the person who inserted the signs referred to, whether the first, second, or a later hand. Not every one of the manuscripts can be assigned with certainty to the century under which it is entered. In doubtful cases I have chosen the latest date. 312 Charles Brewsteb Randolph MAnwscbipts by Centubies* - 5 •oil o a OS u 2 v 36 29 £8 E-i — Thirteenth Centuby (A of Aristophanes -1 P (or A) of Aves 1-200 14 Parisinus 2712 (_ Sophocles 0. C. 1-525 52 11 34 3 26 Fourteenth Century Farnesianus I. E. 5 h of Aeschylus Agam. 1-448 2 Marcianus 468 f... ( B of Aeschylus ( V of Sophocles Agam. 1-348 0. C. 1338-1514 9 16 3 5 1 Laurentianus xxxi. 15 r of Aristophanes Aves 1-200 37 4 31 26 Ambrosianus L 39 M of Aristophanes U of Aristophanes Aves 1-200 Aves 1-200 36 37 28 28 4 Urbinas 141 Laurentianus L of Euripides Hero. Fur. 1-165, 328-480, 607-828, 1166-1219 27 14 7 "G of Euripides Here. Fur. 1-155, 318-462, 625-847, Palatums 287 f { 1143-1224, 1232- 1251 35 11 Vat. of Sophocles 0. C. 1-217 22 13 V Laurentianus L 2 of Sophocles 0. 0. 1-124 14 5 4 Fifteenth Century Laurentianus xxxi. 8 f of Aeschylus ( g of Aeschylus Agam. 1-407 Agam. 1-45, 1095- 10 4 3 Marcianus 616 ) 1673 43 19 b ( V 2 of Sophocles 0. C. 122-319 20 14 b Guelferbytanus . . . b of Aeschylus Agam. 1-310, 1067- 1159;C/M>epft.8-37 26 8 3 Marcianus 222 Aeschylus Agam. 1-310, 1067- 1085 12 4 2 Marcianus 475 G of Aristophanes Aves 1-200 36 3 28 20 Estensis 3 D 8 E of Aristophanes Aves 1-200 37 1 29 19 ♦This table anticipates the second part of my paper in giving also the statistics for the use of the interrogation-mark after sentence-questions, but considerable space is gained by presenting both sets of figures at once. In the case of Urbinas 141, Vaticanus graecus 2291, and Brussels 11343, the Sophocles passage from Marcianus 468, and the Euripides passage from Palatinus 287 the figures are not included in the totals, since the interrogation-point is not found in the parts of the manuscripts upon which the statistics are based. The fac- simile sheets of all these manuscripts except those of Aristophanes were examined in full. t Some parts of these manuscripts contain the interrogation-point, while other parts do not; see the figures in the table. This probably indicates that the work of inserting the sign was begun by a later hand and not finished. These two manuscripts, as well as Parisi- nus 2712 and Marcianus 616, contain plays of more than one dramatist. Sign of Intebbogation in G-beek Manuscbipts 313 Manuscripts by Centuries Fifteenth Century — Continued Palatinus 67 Vp2 of Aristoph- anes Aves 1-200 Parisinus 2886 F of Sophocles 0. C. 1-158 Parisinus 2787 B of Sophocles O. C. 1-172 Parisinus 2711 T of Sophocles 0. C. 1-59 Farnesianus II. F. 34 Sophocles Sophocles O. C. 1-37, 57-76 Marcianus 617 O. C. 1-79 Marcianus 470 Sophocles O. C. 1-83 Vaticanus graecus 2291 Chigi Sophocles Br of Sophocles 0. C. 1-84 Brussels 11343 O. C. 1-139 Sixteenth Century Riccardianu8 34. . Escurial I. Q. 9 . . . R of Sophocles Sophocles O. C. 1-46 ^ntfqr.1241-1353; 0. C. i-212, 318-371 Seventeenth Century Riccardianus 77. . . R 2 of Sophocles Electro, 1490-1510; O. C. 1-8, 661-683 CO a o i| ■2 = m u cq s • o S 8 .3 o H •- 03 ■Si && 33 1 29 16 8 16 9 7 3 5 6 11 6 11 7 11 6 13 7 6 3 35 17 9 5 Of the twenty-seven manuscripts in this list, twenty-four contain the point and comma in combination in the function of interrogation- mark and no other. Summing up the statistics in the case of the individual manuscripts we have the following totals: Times ; is used Word-questions in the 24 manuscripts 577 734 9 Adding to these the figures obtained from the five manuscripts previously cited 11 1,311 20 This is surely sufficient evidence to establish the fact regarding word-questions as I have stated it. The twenty exceptions noted are accidents, and in the case of some of them the reason for the mistake is evident. 1 1 For example, in the passages examined in the Harvard manuscript of the gospels t(s is twice followed by the interrogation-mark (Matt. 7:9, 7:10); but each time a relative clause stands between rls and the concluding words of the sentence, which begin with /*i} in the sense of num., a word regularly followed by the sign of interroga- 314 Chables Bbewstek Randolph If the sign of interrogation is used in a manuscript without any function other than that of indicating interrogation, and is not used with word-questions, it must be used with questions of the other sort, sentence-questions, which are introduced by particles like apa, oil, p,r\, fi&v, or have no interrogative particle, like Aves 1: 'OpOrjv KeXeveK, y rb BevSpov (paiverai ; — "Straight on, you say, where the tree appears?" Now we cannot expect to find any such uni- formity in the employment of the sign after questions of this sort as in the case of word-questions, for two good reasons: (1) a great proportion of these sentences may be read either as question or statement, e.g., the verse just quoted: "Do you say go straight ahead, where the tree appears?" or "You say go straight ahead, where the tree appears"; (2) anyone, whether he knows much or little Greek, can recognize a question-word like rk or irov or ttw, but he must be able to understand the sentence in the majority of cases in order to punctuate a sentence-question correctly. This we know the scribes often could not do. To return to our manuscripts. The first five cited, which I examined at greater length, give us the following information in regard to the use of the sign with sentence questions : Manuscript Symbol and Author Parts examined Sentence- Questions Times ; occurs Ravennas Marcianus 474 Laurentianus ) xxxii. 9 ) Ox. Clark 39 Harv. Gospels Rof Aristophanes V of Aristophanes M of Aeschylus L of Sophocles B of Plato Aves entire Aves entire Prom. Vinet., Agam. Antig., 0. C. 1-525 Euthyphro, Lysis About 80 MS pages 142 140 38 90 208 36 65 91 26 60 104 33 Totals 654 379 Pursuing the investigation through the twenty-four manuscripts of the twenty-seven of which shorter facsimiles were accessible to me, 1 we find: tion. In such instances, indeed, one would probably be justified in classing the ques- tion as a sentence-question, in which event it would not constitute an exception to the rule at all. !The figures in the individual manuscripts are shown in the third and fourth columns of the table on pp. 312 f . Sign or Interrogation in Greek Manuscripts 315 Times ; occurs Per- centage 359 654 188 379 Adding to this the figures obtained from the 1,013 567 56 We may therefore confidently say that the scribes or other per- sons who punctuated a large proportion of our Greek minuscule manuscripts intended to employ the sign of interrogation after what we call sentence-questions, but not after word-questions, or in other words that the sign of interrogation was used where no word in the sentence would have made it clear to the reader that a question was intended. There is of course no way to tell on this side the water how large a percentage of existing manuscripts exhibit the application of this rule, nor would it, I think, be worth while for one person to extend the investigation much farther. It will be very easy for anyone beginning work on a manuscript with whose punctuation he is unfamiliar to test the application of the rule for himself. In the course of this investigation I have examined all the minuscule facsimiles, complete and partial, that have been at my disposal, thirty-four in all. In none of these thirty-four did I find the sign of interrogation in use in a way that contradicts the rule, and in twenty-nine of them I found the sign that is identical with our semicolon, or a slight modification of it — the use of two points and a comma, for example — in use according to the rule. The other five manuscripts, which do not contain the semicolon with interrogative value, are Venetus A of the Iliad Urbinas 141 of Aristophanes (XIV) Brussels 11343 of Sophocles (XV) Vaticanus graecus 2291 of Sophocles (XV) Parisinus 2934 of Demosthenes (X) The manuscript of the Iliad (Venetus A) is punctuated simply by points, and no effort is made to distinguish questions from state- 316 Charles Brewster Randolph ments. The facsimiles of two others of the five just cited show the comma in use in interrogations according to the rule: 1 Manuscript Author Parts examined Word- Questions Times , occurs after questions Sentence- Questions Times , occurs after questions Brussels 11343 Urbinas 141.. Sophocles Aristophanes O.C. 1-139 Aves 1-200 13 37 2 7 28 7 13 The fourth manuscript, Vaticanus graecus 2291, of Sophocles, gives in the small facsimile some slight evidence of the use of the comma with interrogative value, but pending an examination of a larger portion than the 84 verses of the Harvard facsimile it must be classed with the Iliad manuscript above as wanting the inter- rogative sign. The fifth manuscript, Parisinus 2934, 2 of Demosthenes, is one in which the use of the semicolon caused me, like Tischendorf in the examination of its use in a certain uncial manuscript of which I shall speak later, multum negoti. 2 The semicolon not only stands occasionally after word-questions, but also in many places where it is impossible to read any question at all. The solution of the matter, however, becomes clear upon careful examination of the punctuation in different parts of the manuscript. I selected seventy pages for 1 This may, of course, be accidental; there is no way to determine absolutely whether or not the comma was meant for a sign of interrogation in these manuscripts. It certainly is often used in them where it is not interrogative (19 times in the verses contained in the facsimile of the Brussels manuscript, 23 times in the verses examined in U), though this does not preclude its having interrogative value elsewhere; see note 1, p. 309. It is noteworthy that in the Brussels facsimile five of the seven commas used after questions are large, forming a flattened half -circle with the horns in line with the tops and bottoms of the letters without stems, like o and a ; those used else- where in the facsimile are generally smaller. The possibility that the comma may have interrogative value in these manuscripts suggests the question whether the original sign of interrogation in many manuscripts is not really the simple comma, the semicolon arising when a later hand added the comma to a previous dot or dots. Whether the comma alone ever had this value or not, it is certain that the interrogative sign in some manuscripts was formed by a later hand in this way ; for example, in the Oxford Plato : see the description of the punctua- tion in Plate 81, facsimiles of the Paleographical Society ; Gardthausen Griechische Palaeographie, 276; and compare such forms of the question-mark as ,• ., ,. ,: occur- ring in the published phototypic facsimiles (examples in Antigone 47, 449, 510, 734, 1048, in L ; Prometheus 378, 835, 972 in M). ^Monumenta Sacra Inedita, V, xvii. Sign of Interrogation in Greek Manuscripts 317 this, the first twelve and a half (containing the first two Olynthiacs) , and others at intervals throughout. It soon became evident that the occurrence of the semicolon in the manuscript is closely con- nected with that of the comma. At the beginning of the manu- script they are both very common; further over they are both rarer; finally they seem both to be given up entirely. The following table shows the number of times that commas and semicolons occur in these seventy pages: CO 9 is 33 10 4> U OS in So WO S.S GW C M «.5 w-g o c — p. .8 8 !l a , and point in form resem- bling ; to F ^ «4-< O •2»> z5 Olynthiac 1 1 5 18 29 157v 389 6 14 5 mi 14 8 3 42 4 36 2 2 13* 5 1 4 17 2 15 5 3 20 Olynthiac2 24 De Chersoneso 4 4 3 In Cononem Totals 70 93 4 19 23 23 55 *In two of these instances it is possible to read a question. My conclusion is this: The original scribe, and perhaps the first corrector, used the high, middle, and low point. Then came a later hand using the comma. He began to insert it very frequently, as we see from the record in Olynthiacs 1 and 2; then abandoned his original plan, inserted fewer commas, and perhaps finally left off altogether: my examination of the manuscript was not extensive enough to make me certain about this. He did not confine himself to places where there was no mark of punctuation, but freely added the comma where the high, middle, and low point already stood: this is proven by the difference in color of ink of point and comma in many places, and by numerous combinations such as ., ,. , * where the point and comma are not in the same vertical line, where they are in line horizontally, etc. Where the comma was thus freely added, it was inevitable that the semicolon should result in many places, and a few times this has occurred where a question ends. The semicolon in the manuscript therefore does not have 318 Charles Brewster Kandolph interrogative value; it is simply an accidental combination of point and comma. This disposes of all the minuscule manuscripts to which I have had access. None of them shows any contradiction to the rule stated. The use of a question-mark in uncial manuscripts is very rare, and it may be doubted whether in any of those which exhibit it it dates from the first hand. I have found good evidence of its occurrence in only two uncial manuscripts. 1 Tischendorf, in his Monumenta Sacra Inedita, 2 gives transcripts of two biblical manuscripts in uncials dating from the eighth or ninth century, in which the semicolon occurs as a sign of interro- gation. It is in one 3 of these that he says that the semicolon made him multum negoti: that although in many places it is intended to denote a question, it is also found where there is none. The tran- script contains 364 pages, 131 of which I examined very carefully. 4 The examination yielded the following figures: ; used after Word-questions Sentence-questions . 69 71 4 17 The figures given show that the manuscript is imperfectly punc- tuated; 5 the punctuator does not seem to intend to use the sign of interrogation after word-questions. The 131 pages contain the semicolon in no instance where the words which it follows may not 1 The author of the preface to the facsimile of the Fragments of the Illustrated Iliad (Homeri Iliadis Pictae Fragmenta) remarks on p. 18 that he has noted the comma in the manuscript, and also the comma with the point above it, forming ;. I hare been unable upon careful examination of the text to recognize any trace of the semicolon in the facsimile. 2 Nova Collectio, Vols. II and V. zjSpistulae Pauli et Catholicae Palimpsestae, in Vol. V. 4 Time failed me for completing the examination, which, however, hardly promised to be worth while, as any attempt to account for possible irregularities must be based upon evidence which a transcript cannot furnish. 6 The problem of the punctuator in uncial manuscripts was much more difficult, because he lacked the accents to show him whether words like tis, irov, opo, ?!, were interrogative or not. Sign of Interrogation in Greek Manuscripts 319 be read as a question, and I am inclined to think that the incon- sistencies which troubled Tischendorf were accidental. Mistakes are as easy to make in punctuation as in the use of words and letters. In the other manuscript, 1 the transcript of which I examined entire, the use of the sign of interrogation does not conform to the rule for minuscule manuscripts. Without seeing at least a facsimile, it is idle to speculate as to the reason why. Possibly the interrogation marks were inserted after the present practice of ending all questions with the sign became established. When this was may perhaps be determined by an examination of early Greek printed texts, for the earlier books exhibit to some extent at least the punctuation that I have shown to exist in manuscripts. Clark College Worcester, Mass. 1 The Bodleian Genesis, in Vol. II.