Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in
American Journal of Philology
Vol. 48, No. 1 (Jan.‑Mar. 1927), pp1‑9.

The text is in the public domain.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

 p1  Marks of Quantity in the Monumentum Antiochenum

In a paper published in the Proceedings of the American Philosophical Society, LXI (1922), pp87 ff., I considered "The Use of Devices for Indicating Vowel Length in Latin" in the Monumentum Ancyranum, as well as in some other inscriptions. The new Monumentum Antiochenum of the Res Gestae Divi Augusti, published by Dr. Robinson in A. J. P. XLVII, pp1 ff., offers some interesting new material.

Before turning to this new material, a few words about the terminology of one of these devices would seem to be in place, since the usage of scholars differs widely. As to the I longa, used to designate long i, there seems to be general agreement. The mark used to indicate long quantity in the other vowels usually has the form of an acute accent. It was for a time supposed to be an accent, and although that idea was soon given up, the term "accent" is still applied to it by scholars who are perfectly well aware of its real purpose. By Latin epigraphists it is ordinarily called an "apex," a name which has the authority of Quintilian, Isidore, and several of the Roman grammarians. The only objection to the use of that term — if it be an objection — is that "apex" is applied by the Greek epigraphists to quite a different thing. Thus the Roberts-Gardner Introduction to Greek Epigraphy, under Inscriptions of Attica, pp. xv ff., says of certain letters: "The open ends of strokes and the angles where two strokes join are adorned with what are called apices, that is to say, minute cuts set at an angle of 45° to the main strokes, usually one on each side; where two  p2 strokes meet, the apices sometimes take the form of a prolongation of each of the strokes." There are numerous examples of a similar use of the term by Greek epigraphists; in fact, it may be said to be a regular usage.

Important as these ornaments (if we may call them so) often are in determining the date of an inscription, they seem to have received but scant attention from Latin epigraphists. Ricci, Epigrafia Latina, p57, mentions them in these words: "Verso la fine del IIo secolo e durante il IIIo le lettere subiscono la forma che loro vuol dare talora l'artifice, e appaiono o troppo lunghe o con apici alle estremità." On p45 Ricci uses the word "apex" also to designate the mark over a long vowel. Sandys in his Latin Epigraphy has "apex" in his Index with a reference only to its use as a designation of quantity, and apparently makes no mention at all of the "ornaments" referred to above. Egbert, Latin Inscriptions, p63, in describing the different forms of the letter N, says that it was "at first unornamented, afterwards having partial ornamentation, and finally. . . becoming fully ornamented." Here he is obviously referring to what the Greek epigraphists call apices but he does not use that term. In describing the letter M he again uses the indefinite expression "ornamented," but also (p62) calls such ornaments cornua. The equivalent of cornua (κεραία) occurs in Greek as a designation of the so‑called "apex," and cornua itself is used by Isidore (Orig. XVI.27.5), as well as in three glosses, de extremis partibus litterarum; see T. L. L. IV.970.57. The palaeographers frequently use the term "finials" for such ornaments, and either that name or cornua might well be generally adopted, both to avoid confusion, and also because "apex" (originally "top") is not properly applied to marks at the ends of horizontal lines, or at the bottom of vertical ones. But the attempt to reform grammatical nomenclature is notoriously a difficult matter.

In Latin the word apex, as applied to letters, is first used of the mark which indicates a long vowel (Quint. I.7.2‑3). It is described accurately enough, in accordance with the original meaning of "top" or "summit," by Isidore (Orig. I.4.18), as follows: apicem dictum pro eo, quod longe sit a pedibus et in cacumine litterarum apponitur. Est enim littera iacens super  p3 litteram aequaliter ducta. While Isidore calls the apex a separate "letter," it is possible, if not probable, that at first it was joined to the letter over which it stood. It is so found occasionally in inscriptions, and one such instance occurs in the Frag. Ant.: Románum, in the Heading, § 2. While in the strict sense of the word this might be termed a "true apex," it is in all probability accidental; at any rate such cases are extremely rare.

Quintilian (see also I.4.10), Isidore, and Terentius Scaurus (De Orthog., VII.33.5 f., K.) all define the apex as a means of designating long vowels, although the two former incorrectly speak of long syllables; and all agree that its purpose is to distinguish words which differ only in the quantity of one or more vowels, such as populus and pōpulus, malus and mālus, and the like. The T. L. L.​a rightly puts under this caption Quint. I.4.10, where apice has that meaning rather than "circumflex" (L. C. L. I p159). But it is wrong in adding Quint. I.5.23, since there Quintilian is speaking of accent, rather than of quantity. Moreover, the text is corrupt and the reading apice is questioned.

As Isidore uses apex of the mark of quantity and at the same time calls it littera, so the word is used of the strokes which make up a letter, then of a form in general, and finally as the equivalent of littera. When it has the last named meaning, it almost always, if not invariably, designates letters with special reference to their form; for example, Macr., Somn. Scip. I.6.70, septem vocales literae a natura dicuntur inventae, licet Latinitas easdem modo breves pronuntiando quinque pro septem tenere maluerit. Apud quos tamen, si sonos vocalium non apices numeraveris, similiter septem sunt. Here the sound of the letters is contrasted with their form. See also Apul., Metam. XI.22 (end), and especially Heges. V.34.2, impressi illi ante fores templi apices elementorum (= γράμματα). The simple meaning "letter," rather than the form of the letter, is not common; it is perhaps illustrated by the verses of Ausonius discussed below, and it seems certain in Auson., Orat. ad Grat. XVI.74, et adhuc obnoxii in paginis concrematis ductus apicum . . . cernebant; cf. Quint. X.2.2, sic litterarum ductus, ut scribendi fiat usus, pueri sequuntur, and for the meaning "the  p4 form of the letters" rather than "the lines of the lettering" (L. C. L. II, p263), Cic., Fin. V.47, nihil interest. . . qui ductus oris, qui vultus in quoque sit. The meaning "letter" is indicated by many of the metaphorical uses of the word, such as Ulpian, Dig. XVII.1.29.4, de apicibus iuris disputare, and Hieron., contra Ioann. 3, ne punctum quidem et apicem calumniae transeas. It is made certain by the third meaning of the word; for just as the plural litterae means an epistle, so apices has that meaning; for example, in Sidon., Epist. IV.5.1, apicum meorum gerulus, Codex Theod. XVI.2.7, lectores divinorum apicum, and in many other instances, especially in the ecclesiastical writers.

I believe that these three meanings (a mark of quantity; stroke, form, letter; epistle) cover all the examples cited in the T. L. L., and the meanings "stroke" and "form" make some difficult passages perfectly clear. The meaning "summa pars litterarum (litterae)", which is given in several glosses, and which the T. L. L. seems to refer in its literal sense to one or more passages, I believe to refer to the mark of quantity. In Auson., Epit. Heroum, XXXII, we read

Una quidem geminis fulget set dissita punctis

Littera, praenomen sic (·L·) nota sola facit.

Post ·M· incisum est, puto sic (
[image ALT: The letter M with the upper right-hand corner erased.]
	) non tota videtur;

Dissiluit saxi fragmine laesus apex.

The L. C. L. translation (I p159) renders the last line thus; "for the broken top is flaked away where the stone is cracked." But the passage from Ausonius which is quoted above suggests the meaning "letter" for apex, which is used to avoid the repetition of littera, or perhaps for metrical convenience. Dissiluit too is an odd word for "flaked off." I would suggest: "the letter is split apart by a crack in the stone," which would be represented graphically by 
[image ALT: The letter M with the midsection (where the two diagonal strokes join) erased.]

The following rather difficult passages become clear, if we give apex the meaning "stroke" or "form" ("strokes"). In Cassiod., Gramm. VII.148.6, K., digamma nominatur qui duos apices ex gamma littera habere videatur, the meaning is that the digamma has two horizontal strokes in place of single one of gamma. The T. L. L. apparently puts this example under  p5 "summa pars litterarum," but the second stroke of the digamma could hardly be called a "top," even if the upper stroke could be designated by such a term. So too in Ter. Maur. 986, VI p351, K, vel priores G Latini nondum ab apice finxerant, apice surely does not mean the top of the letter. It probably means the stroke which differentiated G from C, or with less unusual syntax ab apice might mean "as regards its form"; cf. Suet. Tib. LXVIII.1, latus ab umeris et pectore. Gellius has two passages in which apices means the "strokes" or "form" of letters. In XVII.9.12, describing the effect of the Spartan σκυτάλη, he says: resolutio autem lori litteras truncas atue mutilas reddebat membraque earum et apices in partes diversissimas spargebat. In the light of Gellius' habit of using pairs of synonymous, or nearly synonymous, words, it seems most natural to translate membra earum et apices by "the parts of the letters and the strokes of which they were formed." In XIII.31.10, where an ignorant grammarian pretended to have worn out his eyes in nightly studies, he says: vix ipsos litterarum apices potui comprehendere, which seems to mean: "I could hardly make out the forms of the letters (die Züge, Weiss)."

Turning now to the Res Gestae, the two versions of which will be designated by Anc. and Ant., we find in Ant. 31 words and parts of words not preserved in Anc., in which long vowels are designated by the apex or by I longa.​1 As one of the words has two apices which are not found in Anc., the total number of marks under this head is 32.​2 There are two errors. Vicíens is probably a mere slip for víciens, which occurs in VI.22; in Anc. viciens occurs twice without a mark. Devíctas is less easily  p6 explained; victor occurs twice in Anc. without a mark, and victoriis in both Anc. and Ant. with no mark on the i.

Excluding three doubtful cases, there are 20 words and parts of words, with 21 marks, occurring in both Anc. and Ant., in which the mark of quantity is found only in Ant.​3 There is one error, omniúm; the word stands between templís and civitátium. The new fragment thus supplies in all 53 new marks.

There are 76 words and parts of words, with 78 marks, in which there is agreement between Anc. and Ant.​4 The mark on ín is probably an error, although fínés follows; a marked vowel before nf is rare, though fairly common before ns. The mark is perhaps intentional, since it occurs in both Anc. and Ant., which is not true of the errors already noted, and to be noted below.

 p7  Over against this list may be set one of 9 words preserved in both Anc. and Ant. which have marks only in Anc.​5 There are besides 8 instances in which a mark is preserved in Anc., where in Ant. the part of the stone which would have contained the mark is broken away or very badly worn,​6 and there are 4 words in which the marking in one or the other version is uncertain.​7 If to the certain cases in which the mark is found in Anc. but not in Ant. we add the undoubted instances in which the reverse is true, we have 30 marks in which the two versions disagree, and 78 in which they agree. This seems, on the whole, to indicate a fair degree of correspondence with the original inscription in Rome on the part of the two versions, but at the same time it is evident that a certain percentage of uncertainty must be taken into account. It may be added that all the editions of the Res Gestae show some variations in their markings, and apparently some errors or oversights. It is highly probable that the stone-cutters of Anc. and Ant. made oversights, and that the original document at Rome had more marks than either copy. For the same reason it is difficult to conjecture whether the same copy was used in Anc. and in Ant. At the same time it is true that the conclusions which I reached in my paper are not affected by the new material that has come to light.

As to the use of the apex, the rule given by Quintilian and other Roman writers is seldom exemplified in the inscriptions, because Latin words which are alike in spelling, but differ in the quantity of one or more vowels, are relatively rare; but á in the abl. of the first declension, and ís in plural cases are fairly frequent, as will appear from our brief lists, to which many  p8 instances might be added. That there were other reasons for using the marks is obvious, and for some of these I must refer to my earlier paper. My confidence in some of the categories was increased by Professor Pease's paper in the Harvard Studies in Class. Phil., vol. XXI, which I had unfortunately over­looked when I published my paper. In his manuscript (see p52) he finds the apex (or "accent," as he calls it) "over exclamatory o." This does not occur in the Res Gestae, and is not common in inscriptions. I have noted one example in CIL VI.5075 (8173). Marks are, however, rather frequent over monosyllables, and in particular over the preposition a. He found marks also "very frequently in compound words, to show that the prefix is not a complete word and carry the reader forward to what follows." In inscriptions we find this sometimes in compounds, such as undéviginti and quinquáginta, and more often in derivatives such as nómen, testámentum, aerárium, and the like. His third category is "to emphasize a long vowel or accented syllable even where there appears little danger of confusion." So far as the inscriptions are concerned, this would have to be emended by the omission of "or accented syllable," since the marks in inscriptions, except for very few errors and some special uses of the I longa, are used only to designate long vowels; also by the substitution of "word" for "long vowel," as in personal names, official titles, and the like. Finally, he notes the mark "on a few unaccented syllables, such as tantó, quantó and pauló." In inscriptions the mark appears quite as frequently over unaccented as over accented syllables, and is quite common over a final o. It is obvious that the use of such marks in manuscripts and in inscriptions differs, and a comparison of the two would be interesting, as well as of different manuscripts and inscriptions with one another. I hope to complete a collection of the inscriptional material within a reasonable time.

As the result of further study and larger collections of material I am inclined to lay more stress than before on the use of the marks as a guide to the correct pronunciation of words. This would account for the fairly frequent marking of vowels which do not receive the accent, but of which the length is important for proper pronunciation, such as tribúniciae, auctóritate, cónsulto, curátionem, úniversi, etc. It would also account to some extent for the inconsistency in the use of the marks, due to varying opinions as to the importance of the length of particular vowels or the likelihood of their mispronunciation. This, of course, applies only to the inscriptions of the educated, in which such marks are most common; with the uneducated they were doubtless conventional or imitative, and used more or less at random; and there were also other reasons for the marks, for which I must refer to my earlier paper. Finally, it is probable that Quintilian and the grammarians had in mind the use in manuscripts, rather than in inscriptions.8

John C. Rolfe.

University of Pennsylvania.

The Author's Notes:

1 For convenience the cases of long i are printed with the apex (í), but this invariably represents an I longa. I am indebted to Dr. Robinson for numerous notes based upon his first-hand knowledge of both Ant. and Anc.

2 The examples are: Heading2, Románum; i.1, á, vindicáví (á found also in Anc.), ; ii.4, trís, vicíens, terrá, immortálibus, diés, tribúniciae; ii.5, á, penuriá, perpaucós; iii.7, reí, Arvális; iii.9, vóta; iii.10, nómen; iv.12, virís, Hispániá (final á found also in Anc.), iís; iv.13, victoriís; v.16, praedís, numeráto; vi.20, auctoritáte; vii.26, nón; ví.30, devíctas; ix.32, liberórum; ix.34, caussá, auctóritáte, céteri; ix.35, cúria.

3 The examples are: ii.5, curátiónem; iii.8, fécí (é is doubtful in Anc. and is not counted); v.15, míllia three times; v.16, próvinciálibus (ó also in Anc.), coloniás; v.17, eós; vi.19, díví (first í also in Anc., Capitólio; vi.20, perfécí (é also in Anc.); vi.21, nómine, consacráví (á also in Anc.; Robinson now reads consacráví), míllia; vi.22, víciens, virórum, vii.24, omniúm; viii.27, máluí (á is doubtful in Anc., and is not given by some of the editors), ix.32, suórum; Summary, 4, mótu.

4 The examples are: i.1, vindicáví (í only in Ant.); i.3, tóto; ii.4, triumphávi (á is not given by the editors of Anc., but it probably should be), imperátor, deposuí, regés; iii.7, sacris (í is not given by the editors of Anc., but seems to be indicated on the stone); iii.8, patriciórum, égí (é not preserved in Ant.), quó, Romanórum; iii.9, vivó, múnicipatim, úniversi, pulvínária (í doubtful in Anc. and not counted); iv.11, senátus; iv.12, Hispániá (first á not preserved in Anc.), áram; iv.14, consiliís; iv.15, congiárium, colonís, sexagenós, plebeí; v.16, agrís, solví, múnicipís (‑nicipis not preserved in Ant.), agrís, próvinciálibus (á probably only in Ant.), deduxérunt, aetátis, quós, stipendís; vi.17, pecunía, meá, iuví, quí; vi.19, , díví (final í only in Ant.), appellári, nómine, Feretrí, summá, sacrá, viá; vi.20, rívos, vetustáte, á, perfécí (í only in Ant.), ampliáto, nón, senátus; vi.21, Mártis, manibiís, consacráví (í only in Ant.), coronárí (both marks); vi.22, dedí, nómine; vii.24, , honórem; vii.25, pacávi, á, sacerdotés; vii.26, Hadriánó (a not preserved in Ant.), flúminis; viii.26, ín; viii.27, máluí (í is found only in Ant.; á is somewhat doubtful in Anc. and is not given by some of the editors), descíscentem, Gáium (the editors of Anc. do not give á, but it seems to be indicated in Mommsen's facsimile), reciperáví (both á and í; Robinson now reads reciperáví); viii.29, aliós; ix.32, suós; ix.33, accéperunt, Médí (é in both; í only in Anc.); ix.34, meó, fíxa; Summary, 4, cúriam.

5 As given in Anc. these are: i.3, remísi; vi.20, reféci; vi.21, féci, appellátus; vii.23, pedés; viii.27, rége, régió (‑io is not preserved in Ant.); ix.33, Médí (i only in Anc.); Summary, 4, quórum.

6 v.16, posteá; vi.19, Mátris; vi.20, plúribus, Márcia; vii.24, meó; vii.26, provinciás, regióne; ix.34, civíca, an error, which in all probability would not have been made in Ant.

7 ii.4, cónsulto, Ant. (apparently no mark in Anc., but the stone is somewhat worn); v.15, vigintí, Ant. (the í is not clear in Anc.); vi.19, Iunónis, Ant. (The editors of Anc. give Iúnonis, but the marks are not clear, and Iúnónis or Iunónis are possibilities); viii.30, imperio, Ant. (the editors of Anc. give í, but Robinson questions the marking. In Ant. he now reads imperio).

8 [The publication of the Journal was too far advanced to permit of the revision of the present article in the light of the new material afforded by Ramsay and von Premerstein's recent edition of the Monumentum Antiochenum, Klio, 19. Beiheft, Leipzig, 1927. — Ed.]

Thayer's Note:

a The most recent version of the TLL is online here.

[image ALT: Valid HTML 4.01.]

Page updated: 13 Apr 19