Skip to main content

Full text of "On the Use of Certain Verbs of Saying in Plato"

See other formats


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 

Read more about Early Journal Content at 
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 


It is hardly more than eight years since there was published in 
Germany the first scientific effort to trace a development in Plato's 
style. Even in this interval the subject has gained such general 
attention from classical philologists that a fresh investigation in 
this field needs no apology. New contributions to the subject 
have now rather to be judged by the quota of truth they bring 
than by the fairness of their motives. 

The present paper treats of certain forms of verbs of saying. It 
embraces, moreover, only those forms that are employed for the 
special purpose of recalling some previous part of an argument ; 
or, more generally, of citing some former passage in a dialogue. 
In the direct dialogues — where each speaker has his own words 
given in a separate paragraph which is headed by his own name — 
such citations make up the greater part of the verbs of saying. 
In the indirect dialogues, however, where the real dialogue is 
related to some person not present at it, these verbs are very 
frequently used by the narrator, not as citations, but merely to 
show changes of speaker. Since in these indirect dialogues the 
persons who carry on the argument make also references to pre- 
vious passages, it may need a moment's consideration to tell what 
duty a verb of saying performs. There is usually, however, a 
difference in the verb itself. Thus, in the narrative it is generally 
'i$rp> or rjv, more rarely throv ; but the verbs used for reference show 
much more variety : \4ya> exhibits perhaps the most extensive use, 
(faov is more frequent in certain dialogues, 'i^p occurs very rarely 
(Protag. 359 A and B), and rjv never. The real difference turns, how- 
ever, on whether the verb is used for narrative or discussion. Both 
kinds of usage may even be observed in the samespeaker when, after 
narrating a dialogue, he begins to analyze it or to consider its 
importance. An illustration will make the point clearer. Phaedo, 
in the dialogue named after him, is mostly a narrator, and, as 
such, uses the ordinary verbs of narrative, as ?<£>), § 8' &. At 88 
B he turns to discussing with Echecrates the story he has just 
been telling, and then the phrases irpoeipwevats Xdyois, w 6 2o><cpaTJ)s 


e\eye, are citations of the same kind as would be found in a direct 
dialogue. Furthermore, whenever Socrates, Cebes, or Simmias 
refer to their own discussion, they of course use references of this 
same sort. In the Euthydemus, to take another example, refer- 
ences are made by Socrates in his argument with Euthydemus 
and Dionysodorus, and with Crito. On the other hand, in his 
narrative to Crito he uses verbs of saying that are peculiar to the 
indirect dialogue, and, consequently, are not citations. Perhaps 
it is in the participle that the verb of citation is most easily mis- 
taken for that of the narrative ; but the following examples will 
show the difference : Parmen. 135 E,n-Xijv tovto ye aov koI irpos tovtov 

riyao6r)v (lirovros, is a citation ; but Protag. 334 C, dirovros olv ravra 

airov ol irapovres avcdopv^aav, belongs to the narrative. 

When all the references have been collected, they are found to 
make up a considerable number of instances for each dialogue. 
Some of them are in the present tense, as tpr/s, Xeyeis, and they 
then recall some statement that is distinctly before the minds of 
the speakers or has just been uttered. Past tenses, as !Xey<u>, 
f'ppnBil, serve to bring up a statement that is more distant or may 
be just falling out of memory. They are not, therefore, so specially 
suited to passages of short, vivid question and answer, but are 
introduced at some distance from the statement which they refer 
to, and besides are found in all varieties of dialogue. If, then, 
their occurrence is not due to the form of the dialogue, or to any 
other accidental circumstance, any changes that occur in them 
must be due to alterations in Plato's style. On examining all the 
references which are made by past tenses of verbs of saying, it is 
found that some dialogues show an unusual variety in the tenses 
of the passive. By means of this increase in variety the dialogues 
can be arranged in a series which, it is intended, should show 
primarily the course of Plato's development in style. This order 
is given in the annexed table. 

In the first place, the statistics on which the series is based 
should be explained. In determining the frequency of any given 
form of citation in such different dialogues as the Symposium, 
Gorgias, and Timaeus, it is evident that a standard, such as the num- 
ber of pages covered by the dialogue, will not give a just relative 
proportion. A truer measure seems to present itself in the total 
number of references found in each dialogue, and these totals are 
what is given in the first column of the table. They are composed 
solely of references to the argument, and do not include any except 


b o g-s-o s-S p-^ 

w Cfl to S. K"g S. X'co w r* a- a> ^ 

£T " tft ,- " CU rn t-t li 

'.' - g- 

ro ^< rm "^ ° 

W 3 td' 

1 < 

'u OOOJ VO 0%-f* OOW S-fe'^.O"^ " M X" " M M " 

rt 1-3 

8> O 3 VO 3 00 O ChCn M m g> j^S. 2, o 

m 4^ -<j io bw m *o b on bo : : ::'-:::::: : ?" -s 1 s 

OJ 10 

avw 00 o ONCn^-f^ m k> u M ::::::::::: : l-a^-s" 

• . -^ SOS 3 

ti 2 /-»>n ^t ^, r.> Jk w 1-1 n ^ ki « 1- „ m M ; : : : : • 5". 5-^4. ~> X f" 

0v 1-1 fJiVO «-l hi W -*» wmOmM 


5. s, S o P 

tnMw-uwooo^-ikOojMM: : : : : : : : ; : 2. „ -g-g £ ^ 

'TJ ETC* "^ ^ 

tt 2. 3 "3= 2. 
ui ^ s; ^ -*i 

O <->\ hi K) w 

• P 

S t> , 


->p S 2. O 

yw — 

9 « a 

w h n ; '. : £t£.~, 

? S 5. » « 

p ~J 

[ -.H'-o H 3 v) w h m o ti 2^P t) r r m > 

" C 3 " £ 9 " 

01 3 3 w 



such as go back to full, complete statements of the persons engaged 
in the discussion. Thus, all words are omitted which refer to 
quotations of poetry, of tradition, or of any composition not 
original to the speakers of the dialogue. The only exception is 
the oration of Lysias in the Phaedrus. This is regarded as a 
component part of the dialogue because it is read in full by Phae- 
drus and plays a prominent part in the succeeding discussion. 
In the whole argument Phaedrus himself acts really as a passive 
listener, and the opinions considered belong either to Socrates or 
Lysias. In contrast to the speech of Lysias are such passages as 
the poem of Simonides in the Protagoras, the story of Atlantis, 
the myths of Er and Gyges, and various quotations from Homer 
and others taken up by Plato from without. These are not con- 
sidered essential parts of the real dialogue, and citations of them, 

like ravrd re otv navra wpbs rbv XIittokov eipryrai, or ra piv Srj prjde'vra bird 
tov naXatov Kpnlov Kar aKorjv rrjv SoXtovo? aKrjKoas, are not Counted. 

Furthermore, in explanation of this first column of the table it 
should be stated that all the citations summed up in it are made 
by past tenses of the verbs Aeya>, ipS>, ehov, and <prjpi. Most of 
them are in the indicative mood, as eXeyov, upryrm, eX^V ; others 
are past participles, as ewrw, to. prjBhra, to dpr/peva. All these 
words must refer to statements already made, and cannot there- 
fore be qualified by the negative or the particle av. Excluded 
from the table are also all infinitives and imperatives. Past tenses 
of these moods, indeed, often refer to previous statements, but 
they do not do so necessarily ; and, at times, it is difficult to decide 
what their exact effect is, so that, on the whole, it has been found 
better to omit them. Their various uses may be seen from the 
following examples taken at random : Soph. 258 E, ml navranao-l 
ye dXij&orara pot SoKoCpev elprjicevai, this certainly refers to the past, 

but 222 B, roirav onorep' av 17 (plXov (IpijcrBal <roi, tovto fip.1v biopiariov, 

refers to the future, or, at least, is not a citation of any preceding 

passage. Laws 737 A, (lpfjO-6a> Sh, vvv 6Vt hib. tov pf/ (piKoxpripareiv 

k. t. X., 738 B, S« Se avTa prjBijvat, are other examples that are not 
references. So, too, ?&>£<«• \i~yetv is a citation, but evidently a 
line of limit must be drawn somewhere, so for convenience it is 
taken in such a way as to include all past tenses of the indicative 
and all past participles of the four verbs mentioned above. 

The second column of the table shows what percentage of these 
citations is formed by ippfjdrj and fts participle pijtius and the third 
column gives the absolute number of these special forms. The 


fourth column shows the cases of e\e\6r) and Xf^k when used as 
references, and the next does the same for npocpp-qdri, npoeiptjrai and 
their participles. A peculiar and rather harsh construction of 
bezels, as an adjective qualifying a noun of masculine or feminine 

gender, e. g. Phileb. 52 C, rj&ovas . . . an.a8apTovs 6p8S>s hv Xex#«'<rar, 
Soph. 219 C, t«x«") tis ktt)tikt) Xex#eT<ra hv himrptyeitv, is shown in the 
sixth column. Cases of the rare perfect passive of X<-'ya> are given 
in the last column. Some of these, however, are imperatives, and 
it should be remarked that the last two columns are not restricted 
to citations, but include all instances of the forms mentioned. 

With these statistics in view, the next step is to observe how 
they determine the order of the dialogues. The first six do not 
show any of the forms given in the table. These can, therefore, 
only be put into a group by themselves, while their relations to 
one another within it have to be left undetermined. The next 
few dialogues in the column owe their position to the fact that 
they begin to show instances of eXex&n- Then, when ipprjOrj begins 
it is chosen as a criterion, and finally the Xe^e/r-construction shown 
in the sixth column becomes the test-word. Thus the early stages 
of each usage are considered to be most important, as it is then 
that the employment of the special word, is most a matter of con- 
scious effort. When any of the test-words becomes so frequent 
as to be used, say more than four or five times in a dialogue, it 
has evidently become well established in the author's vocabulary, 
and thereafter the number of times it is used will depend rather 
more upon the character of the dialogue and on various " acci- 
dents." Especially is it true that a conscious effort has to be 
made whenever a new synonym is introduced, and these citations 
are all more or less of synonymous meaning. A word used to 
convey a peculiar or novel sense might have long been in an 
author's mind before he would have occasion to use it, but a new 
synonym would seem to be introduced rather for variety or other 
artistic principle, applied as soon as its value was perceived. The 
columns containing 7rpoeppr)dt) and XeXexrai have, in general, a ten- 
dency to confirm the evidence of the others , but, except for this, 
are not of so much importance in fixing the order of the dia- 

After the use of any one of the test-words becomes well estab- 
lished, as in the lower part of the columns containing ipptj8r) and 
e'Xf'x&j, its decrease or increase, as noticed above, may be due to 
the special character of the dialogue itself. The Laws is rather 


deficient in the number of aorist passives of both sorts, and the 
Phaedrus seems to have too many of them to warrant placing it 
before Books VI-X of the Republic. In the Gorgias there is an 
entire lack of iXe'xdr], not to speak of several blank spaces in the 
column of trpoepptfr), and an apparent excess of this same verb in 
the latter half of the Republic. 

Some of these irregularities allow of an explanation. In the 
Phaedrus, for example, the excess of ippi)6r) is merely part of a 
general preference for passive forms which this dialogue shows as 
compared with its immediate neighbors. Thus the Symposium 
has 7 citations in the passive out of a total of 29, Phaedrus has 20 
out of a total of 39, Republic (Bks. VI-X) has only 15 out of a 
total of 80. 

Most of these passives in the Phaedrus refer to the speeches 
of Lysias and Socrates. Now, Lysias is not present to explain 
objections to his proposition, and the words of Socrates are 
uttered under show of wild enthusiasm. In order, apparently, to 
avoid giving too much personal responsibility for the statements 
in these speeches Plato refers to them by a verb in the passive, 
for by so doing he keeps the authors more out of sight. 'Bpprj- 

6t]TT)V to> \6ya>, Xdyof 6 e< tov /3($Xi'ou prjBtis, and in the active o>? ra biovra 

ilprjKOTos tov TToijfrov may be given as instances of this sort of indefi- 
niteness. Among the abundance of passives in the Phaedrus it 
would be only natural to find more than the usual number of 
aorists, as that tense had become quite familiar at this stage of 
Plato's style. In compensation for this tendency to use passives 
the dialogue has therefore been moved one place upward in the 
series given in the table. 

For the case of the Laws a different reason is at hand. 85 of 
its 324 references are made up of etprjrat and its participle, while 
no other dialogue shows one-seventh as many of these forms. 
Their frequent occurrence in the Laws is due to its imitation of 
the precise phraseology of statutes and decrees. As such it can 
be easily paralleled by inscriptions belonging to Plato's lifetime, 

e. g. C. I. A. II 17) w Xueu> n fiei tS>v iv rwiSe tgoi ■^rytpla'paTi elprjpevav 
(54), or II 38, pepicrai 8e to dpyvptov to elprjpivov (19), Or Mittheil. II 

142, 6s elprjTai iv tS>i -^rjtplo-pan (20). Consequently, this legal form 
of citation has the effect of diminishing the instances of ippffirf 
and iXixdr) in this dialogue as compared with the Timaeus and 

II. Thus far the table has only been asked to show a gradual 


change and development in Plato's style. If this is once granted, 
another argument, already made use of by Dittenberger, will lead 
to the conclusion that this series of the dialogues has a chronolog- 
ical meaning also. Some dialogues at the upper end of the list 
are known by tradition to be early and some at the lower end are 
known to be late. Hence the most natural course for any one 
who has advanced thus far is to conclude that the table shows 
approximately the order of composition of those other dialogues, 
about which there is no tradition now remaining. 1 

The present paper, as we have seen, treats a number of words 
that are synonymous in so far as they consist of references to pre- 
ceding passages or are peculiar forms of verbs of saying. Xeno- 
phon is so much more historical than argumentative, that the 
number and variety of his citations is too meagre to afford any 
basis' by which to date his works even approximately. Since, then, 
the references shown in the table find no corresponding develop- 
ment in Xenophon, they can best be explained as a result of Plato's 
unusual love of variation, and not as a general change at work on 
all Greek prose at that time. When once taken up by Plato these 
special forms of citation seem to have been employed in gradually 

1 This method of arranging the Platonic dialogues in the order of composi- 
tion by means of changes in their style was first used by Dittenberger, in 
Hermes XVI 321, through statistics of certain phrases containing fiyv. 
These phrases were not strictly synonymous, but their importance in showing 
time of composition was partly because some of them seem to have been 
taken up almost simultaneously by Plato and Xenophon after their early 
writings had been finished. Schanz, who made the second investigation 
(Hermes XXI 439), found that certain synonyms, ovtuq, t& ovti, and dX?/(tof, <jf 
akt)8ag, alijdeia, were first used in but one or two forms, and that at a later 
period the others were introduced and used side by side with them, until at a 
third stage the earlier forms were wholly or partly supplanted by the later 
ones. Statistics of less relevant bearing on the order of the dialogues have 
been published by A. Frederking in Jahn's Jahrb., 1882, p. 534. These show 
the usage of fiov, of certain instances of re, and of el-rrov (when employed for 
narrative, but not as a reference). C. Ritter's comprehensive book (Untersuch- 
ungen uber Plato, Stuttgart, 1888) was unfortunately not accessible during the 
preparation of this paper. A full summary will be found in the present number 
of the A. J. P., so that it will only be necessary to note here in what respects 
the above results are anticipated by his thorough and masterly treatise. The 
value of epiirjBt] as a test-word has not escaped him, although his statistics give 
only the cases where it is used as a reference in a relative clause. His impres- 
sion that ippifin and (Mx® 7 ! are pretty much limited to the last five dialogues in 
the list is, of course, superficial and erroneous. The Charmides, a dialogue 
placed almost at the beginning of his list, contains to ?,ex&ev (162 E). 


increasing numbers, so that the forms previously in use inevitably 
suffered a slight decrease, but were not at any time discarded, as 
in the case of some of the synonyms examined by Schanz. In 
other words, it was a mere question of variety, a growing rich- 
ness in Plato's vocabulary, and not an effort to square with a 
prevailing fashion or attain new shades of meaning. 

The arrangement of the dialogues, as given in the table, 
approaches much more closely to that of Dittenberger than that of 
Schanz. The short dialogue Lysis receives in the present scheme 
a somewhat earlier position than it has with Dittenberger, and is 
not so surrounded by dialogues larger than itself. The Parmenides 
is also placed here rather earlier, so that the suspicions which are 
current in regard to the authenticity of these two dialogues seem 
somewhat strengthened. Greater differences are found on com- 
paring the present order with that of Schanz. In his arrangement 
the Euthydemus and Theaetetus are brought close together, and 
both are placed before the Republic. Moreover, the perplexing 
Phaedrus is put before the Cratylus and Euthydemus, although 
it is not easy to see how the statistics of Schanz afford any proof 
for such an early position. 

The primary object of the above statistics is to assist in estab- 
lishing an unquestioned value for the stylistic method of finding 
the" order of the dialogues. When, by accumulated evidence, 
this has once been done, the second step should be to compare 
its results with those reached by dating the dialogues from their 
philosophic contents. If the two paths coincide at certain points, 
these places may be regarded as definitely settled ; but where they 
diverge, of course an estimate of their relative accuracy must be 
attempted. As yet the results obtained by statistics in regard to 
style are too few and, in some minor points, too discordant with 
each other to call for careful comparison with such facts as have 
been established by the older system of investigation. In this 
contribution, therefore, all reference to the philosophic content of 
the dialogues, and to the various and opposing theories as to the 
development of Plato's philosophy, has been purposely omitted. 

George B. Hussey.