"Peri Hypsous" [Greek] and Aristotle's Theory of the Mean Coulter, James A Greek, Roman and Byzantine Studies; Fall 1964; 5, 3; ProQuest pg. 197

IIEpl ery woue; 3.3-4 and Aristotle's 
Theory of the Mean 

James A. Coulter 

I N A RECENT STUDY of "tradition and originality" in the anonymous 
treatise On the Sublime, W. Buhler1 has demonstrated in a manner 
far more conclusive and thorough than heretofore the full measure 

of "Longinus'''2 debt to the rhetorical and literary-critical tradition. 
Absent, however, in Buhler's monograph, although this is admittedly 
a selective commentary, and in the scholarly literature in general, is 
a satisfactory account of Longinus' relation to Aristotle and to the 
Peripatetics. One factor which has probably tended to divert attention 
from this question is the overwhelmingly Platonic character of the 
treatise, however we may disentangle the line of inspiration which ran 
from Plato to Longinus.3 Nevertheless, the influence of Aristotle, and 
Theophrastus too, on later literary theory appears to have been im­
mense, a fact which is becoming increasingly more evident with the 
advance of scholarly work in this field.4 The a priori likelihood, there­
fore, of some debt to Peripatetic stylistic theory is considerable and 
invites a more detailed investigation. 

Scholars have, it is true, occasionally noted Peripatetic elements in 
the treatise, although, for the most part, either in an unsystematic 
manner5 or merely in passing,6 without any attempt to assess their 

1 Winfried Buhler, Beitriige znr Erkliirung der Schrift vom Erhabenen (Gottingen 1964). 
Z The use of the name "Longinus," hereafter without quotation marks, is meant to 

imply nothing concerning either the identity or the date of this figure. The only con­
sideration has been convenience. 

3 Posidonius or Theodorus of Gadara, or both together, are the two vehicles of influence 
which have been chiefly suggested. Mutschmann (see below, p. 198 and n.7) was the cham­
pion of the latter. His position, however, and that of those who follow him (e.g. Rostagni) 
has been seriously weakened by Grube's attack (AJP 80 [1959] 337-365). The case for Posi­
donius seems strong; see Aulitzky, "Longinus," RE 13 (1927) 1420.63-1422.49 and Reinhardt, 
"Poseidonios," ibid. 22.1 (1953) 772.26-34. 

4 For Dionysius of Halicarnassus, see S. F. Bonner, "Dionysius of Halicarnassus and the 
Peripatetic Mean of Style," CP (1938) 257-266; for Horace, see C. O. Brink, Horace and the 
Art of Poetry (Cambridge 1963); for Demetrius, see G. M. A. Grube, "An Ancient Literary 
Critic," The Phoenix Supplementary Volumes 4 (Toronto 1961) 32-38. 

5 E.g., K. Swoboda, "0 spise II€p' ·y,pov,," Listy filologicke 52 (1925) 83-85. 
6 E.g., F. Solmsen, "The Aristotelian Tradition in Ancient Rhetoric," AJP 62 (1941) 184. 

Here the question of Peripatetic influence on Longinus is noted, but only briefly. More 

197 



198 IIEP! Y'l'OYJ: AND ARISTOTLE'S MEAN 

significance in a more than casual way. In this particular matter, how­
ever, such treatments are especially unsatisfactory, since they imply 
an estimation of Longinus which contradicts what is most charac­
teristic of this author. For scholarly investigation has shown that 
however certainly we may be able to derive any particular feature in 
the treatise from some predecessor, the clear fact is that Longinus 
cannot on the whole be charged with merely assembling bits and 
pieces of older systems. The key elements in his treatise, drawn though 
they may have been from diverse sources, were welded by this extra­
ordinary critic into a cohesive and original system of aesthetics. In 
view, therefore, of both these considerations, it seems a distinct lack 
that there has been so little attempt to make more sense of these 
observed cases of Peripatetic influence, i.e., to interpret them in the 
light of the general intentions of the treatise. 

I 
In one of the most influential studies of Longinus published in this 

century, the Tendenz, Quellen und Aufbau der Schrift vom Erhabenen,7 
Hermann Mutschmann, although concerned chiefly with elaborating 
his theory of a Theodorean Longinus, paused for several pages (91-94) 
to examine elements in the nEp~ "YifJOvs which appeared to him 
clearly Peripatetic in character. Mutschmann maintained that when 
Longinus at 3.4 (&'u'a TO /LEV oloooll KTA.) gave a name to the vices which 
were akin to the <sublime: viZ" <tumidity' (TO oloovII) and <puerility' 
(TO /LELpaKLwoES), he constructed his analysis on the foundation of the 
Aristotelian theory of the mean. For Longinus, so Mutschmann 
argued, <tumidity' was an excess in the Aristotelian sense, <puerility' 
a deficiency; the third vice, 'IT'aplvOvpaov, was something of an in­
truder, and Mutschmann pointed to several indications which bear 
out this assertion. 

Before considering Mutschmann's evidence-and he is surely right, 
although, as I shall argue, in only a qualified sense-I should like to 
examine the passage which slightly precedes the one upon which 

recently, D. A. Russell ("Longinus" On the Sublime [Oxford 1964]) notes ad 3.4 llitt '7'6 JLEV 
ol8ovv • •• ayO'V£O"TaTov: "The transition ... is managed by means of a scheme based on 
the Peripatetic doctrine of the mean." He does not comment, however, on what he con­
siders the significance of this fact to be. Hanns Stefan Schultz (Der Aufbau deT Schrift II£p~ 
·Yr/Jov!> [Diss. Berlin 1936] 42-46) does however essay an interpretation of elements which 
derive, he argues, from Theophrastus' II£pl. A£g£w!>. 

7 (Berlin 1913). 



JAMES A. COULTER 199 

Mutschmann based his argument and which he did not consider. 
Longinus is analyzing the process by which the author who aims at 
'sublimity' falls instead into the vice of 'tumidity':8 

"'"\ s;,'" l' "s;, ~, ~ '"\ s;,,I,. "\ 
OI\WS" U EO£KEV £tva£ TO O£U£tV EV TOtS" fLal\£(T'ra uvu.,.,Vl\aK-

, ,/.. , ,~ C"O!JA.. ' ,.t.. , 
TOTaTOV • .,.,VUE£ yap a7TaVTES" Ot fLEYE OVS" E.,.,£EfLEVO£, .,.,EVYOV-

, (J I 'i:.' , , .,~, ~ " 
TES" au EV€LaS" Kat s "7pOT7JTOS- KaTayvwuw, OVK OLD 07TWS" E7Tt 

~8" ,I,. , 8 ' ~"\' \ 8 ' TOV V7TO.,.,EpOVTa£, 7TH 0fLEVO£ Tep fLEyal\wv a7TOI\LU aLVELV 
., " t, ) , ~\ N , , , , 

OfLWS- EVyEVES" afLapT7JfLa • KaKOL uE 0YKOL KaL E7TL UWfLaTWV 

, \ , • ~ ", '0 " , Kat I\oywv Ot XaVVOL Ka£ aval\YJ ELS" Kat fLYJ7TOTE 7TEpLLUTaVTES" 

• ~ , , , ''''' , ,I,. 'c' .'" ~ YJfLaS- E£S- TovvavTLDV' OVOEV yap, .,.,au£, s YJPOTEPOV VOpW7TLKOV. 

The concepts employed in this analysis of stylistic aberration, and 
the terms used to express them, strongly suggest that Mutschmann's 
assertion of Aristotelian influence merits attention; for here, too, 
traces of Peripatetic doctrine are surely to be detected. Let us examine 
the process, as described by Longinus, by which the author (1 shall 
point out below the significance of the italicized terms) who aims at 
'grandeur of style: in an effort to escape from the charge of 'dryness' 
and 'insipidity: lapses instead into 'tumidity'-a danger difficult to 
guard against-and even, at times, into its opposite, 'dryness: the very 
extreme he sought in the first place to avoid. Now except for the 
notion, for which I can find no Aristotelian parallel, that overshooting 
the mark occasionally brings us back to the very fault from which we 

h ( , , 'f... , , '3 4) h soug t to escape Kat fLYJ7TOTE 7TEPLLUTCXVTES 7JfLCXS EtS TOVVaVTtov . ,t e 
dependence of this passage on the concepts elaborated by Aristotle in 
his discussion in Books 2 and 4 of the Ethics appears for the following 
reasons highly likely. 

First, in the passage quoted, virtue (now a virtue of style) appears to 
be viewed as a mean which lies between two vicious extremes, into 
either of which an author may easily lapse. That the notion of the 
mean and its two extremes is present in Longinus' discussion of 
'tumidity' is suggested, first of all, by the statement further on in this 
passage [3.4] that "tumidity desires to surpass the sublime." Now, 
inasmuch as the vifn7 of 3.4 is merely synonymous9 with fL'YEf}OS of 3.3, 
it seems clear that in the passage under discussion also the vice of 

8 The citations from the n£p~ ·Y!foov) are from the fourth edition of the Jahn-Vahlen text 
(Leipzig 1910). 

9 Pace R. Weber, Die Begriffe 11-''1£80) und v!foo) (Diss. Marburg 1935). See the damaging 
review of Stegemann, Ph W 59 (1939) 638-643. 



200 nEPI Y¥'OY.E AND ARISTOTLE'S MEAN 

<tumidity' should be viewed as an excess, in very much the sense that 
Aristotle understood this term; i.e. as an attempt gone awry, which 
overdoes <sublimity.' As in the Ethics, it is an imprudent overuse of an 
otherwise commendable quality which engenders vice, this quality 
being, in the present case, a style rendered impressive by the use of 
metaphor. In the same way, it is clear that aU8€V€La and e7JP677}S", 
which our hypothetical author endeavors to flee, are deficiencies in the 
exact Aristotelian sense, since they are at the same time qualities 
which by their nature are opposite to another extreme quality, i.e. 
<tumidity: (TovvaJITlov 3.4) and which ought also to be shunned 
(cpEVyOJITES 3.3) by anyone who aims at <sublimity'; for aU8€V€La and 
~7Jp677}s, to judge from their meaning, are deficient in precisely those 
characteristics of style which Longinus designates as <sublime': they 
are bloodless and feeble. We would thus have the following scheme: 

tnr€pf3o"ATJ - TO ol8ovv 
, , 1I 

JL€U077}S - JL€YEUOS 
;AA€L~LS - aU8€VELaj e7JP6T7Js 

The interpretation implied by this scheme seems the only valid one 
which can be extracted from the passage. <Aridity' is clearly anti­
thetical to <sublimity'; it is otherwise difficult to see why we should 
have to flee it [3.5]. In any case, common sense tells us that this is true. 
We are told, furthermore, that in aiming for <sublimity' we some­
times end up with <tumidity,' which is described as the opposite of 
<aridity.' <Tumidity,' moreover, is seen in some way to be an endeavor 
to exceed the <sublime' [3.4]. In 3.3-4, therefore, Longinus is telling us 
three things: first, that <aridity' and <tumidity' are opposites to each 
other; second, that <sublimity' lies somewhere in the middle between 
these two extremes; third, that of the two extremes, it is <aridity' from 
which those who aim at <sublimity' endeavor to flee. The similarity to 
the terms of the Aristotelian discussion of the mean is striking. 

The terminology, too, is precisely reminiscent of the Ethics. The 
author in Longinus flees (cpEVyOVT€S 3.3) from <aridity,' which is des­
cribed as the opposite (TovvavTlov 3.4) to <tumidity.' So, in the second 
book of the Ethics,I° 7Tfis E7TLUTTJJLWV T1]v V7T€pf30"A~v JLEV Kat. T1]V ;AAEL~W 
../.. 1 [ 0 b5--6] d • \ \" \ ~ 1 \ -'- \ \ 1\ 'f'€vy€t 2.11 6 an , at JLEV yap aKpat Ka£ 77J JLEar! Ka£ U/\I\7Jl\aLS 
€vaJITla£ Elulv [z.l1osb13-14]. The iCPtE{-£EVO£ of3.3 may also be compared 

10 Citations from Aristotle's Ethics are from Bywater's Oxford text (Oxford 1891). 



JAMES A. COULTER 201 

with Ethics 1.1094al-2, 7Taua T'XVTJ . . . &yaOov TtVOS Ecpl€uOat 80K€t; 
although it should be noted that the customary Aristotelian term 
in this connection is UTOX&'€UOat (cf ibid. 2.1109a30 T6v uToxa'&/L€voV 
TOU /Lluov). Again, in discussing the various rules for striking the 
mean, Aristotle notes that in matters of pleasure we ought especially 
to be on our guard (wfAtura cpvAaKr€OV Ethics 2.1109b7); Longinus 
likewise, although perhaps for different reasons,n remarks [3.3] that 
'tumidity' is €v rotS /L&Atura 8vucpvAaKrorarov. 

Mutschmann's assertion of Aristotelian influence on Longinus' dis­
cussion is therefore well founded, although it has to be modified in 
two important ways. First, the specific terms of Mutschmann's analy­
sis are not valid; this is, I hope, already apparent from my own discus­
sion (and see below, p. 205). Second, Mutschmann's remarks suggest 
that he was unaware ofHendrickson's12 thesis that Aristotle employed 
the doctrine of the mean, which he had already elaborated in the 
Ethics, as the central concept in his discussion of stylistic virtue in 
Rhetoric 3, chapters 2 and 12. For Aristotle, and Hendrickson is surely 
right despite the demurrers of several scholars, uacp~v€ta was a 
stylistic virtue (and Aristotle's only one) which was defined not abso­
lutely, but only relatively to the necessity of avoiding, on the one hand, 
utter baldness of manner (i.e. ra7T€£vTJv [Elvat] Rhet. 3.1404b3) and, on 
the other, a style rendered unintelligible by an excess of metaphor 
and gloss (i.e. V7T~p r6 &glw/La Rhet. 3.1404b3-4). Accordingly, Mutsch­
mann's contention that Longinus derived the terms of his discussion 
of 'sublimity' and its vicious extremes directly from Aristotle's treat­
ment in the Ethics is not necessarily correct. The fact is that Aristotle, 
as was demonstrated by Hendrickson, had himself already made this 
adaptation in the Rhetoric. And in this step he was to be followed 

11 In Aristotle the matter is clear. The extreme which is pleasurable is the more difficult 
to guard against, because the pleasant is something to which we are naturally attracted. 
In Longinus there is some uncertainty about why 'tumidity' is especially difficult to escape 
from. To judge from the sentence which follows. it appears that when we aim for 'sub­
limity,' we lay ourseh-es open to a dangerous natural (",VO'(£L) tendency to lapse into 
'tumidity: But there are no clear indications as to why this is naturally so. The probable 
explanation is that for Longinus 'tumidity' is. first of all, the extreme more akin to the 
'sublime,' and therefore at times separated from this quality by only the finest line. And 
given Longinus' views of the natural affinity of the soul for the 'sublime: it is not difficult 
to understand why most blunders should occur in the extreme which is closer by nature 
to 'sublimity: 

11 "The Peripatetic Mean of Style and the Three Characters," AJP 25 (1904) 125-136. 
Hendrickson's view has not gone uncontested. For another analysis, see Solmsen (above, 
n.6) 43. 



202 nEPf YlJfOYlJ AND ARISTOTLE'S MEAN 

frequently in later antiquity,13 a consideration which may help to 
mitigate the seeming improbability of the notion that Longinus was 
in this particular discussion closely indebted to Aristotle. For stated 
as it was by Mutschmann, the idea seemed all too unlikely. Why, 
after all, would Longinus, whose outlook is so thoroughly Platonic, 
turn to the bald details of the Aristotelian system of ethics for his 
terms and concepts? The question however, natural though it seems, 
ignores the two factors just considered: Aristotle's own use of the doc­
trine of the mean in his stylistic writings, and the continued presence 
of this concept in the rhetorical writers of later antiquity. 

To return to our analysis. Having considered TO o18ovv, we may 
examine the vidous extreme which Longinus tells us [3.4] is directly 
opposed to sublimity, viZ" <puerility' (TO p,£LpaKLw8E~) or <frigidity' 
(ifroXp6'T7'J~). (The large question which the "or" of the preceding sen­
tence begs I shall consider presently.) Now the terms in which 
Longinus describes the dangers which lead to this aberration exhibit 
important similarities to the previous discussion of <tumidity: As 
there, the author who strives for (OpEyop,EVOL, cf iCPLEp,EVot 3.3) a certain 
stylistic quality-here the 'pleasant' (TO ?j8v)-lapses (oAul0alvoVGL, cf 
inrocpEpOV'TaL 3.3) instead into the related vice, ·puerility.' The chief 
difference here is that, whereas in the passage on <tumidity,' Longinus 
presented the full scheme of the mean and its two extremes, in the 
present passage he treats the matter only in terms of a given virtue of 
style, here the <pleasant,' and a single related vice; in other words, in 
terms of the doctrine of ap,aprljp,aTa 7TapaKElp,Eva or vicina vitia, which 
is familiar from Demetrius, Auctor ad Herennium and Horace. This 
curtailment, however, should probably not be viewed as a deviation 
from the Peripatetic character of the discussion, since the doctrine of 
viana vitia was itself almost certainly nothing but a simplified deriva­
tive of the theory of the mean as it was applied to style.a 

18 As, for example. by Dionysius of Halicarnassus; cf. Bonner (above. n.4). The "real" 
Longinus, in the third century, himself shows clear traces of the Peripatetic doctrine of the 
mean in his TIXVTJ P"7TOPLK", (Rhet.Gr. ed. Walz. IX. 559 TO 8' lpyov •.• TOV p.lTpov >.ap.{Ja"11 
560). Moreover. it is likely that the wide-spread doctrine of ttp.apT"'p.aTa 'TTapaK€lw:va was 
itself merely a simplified derivative of this doctrine; cf. below, n.14. 

a This appears as likely a solution as any. It is as if. for example. Aristotle's discussion of 
p.eyaAo1TpI'TT€La in Book 4 of the Ethics had been Simplified by some successor. Instead of a 
scheme with one vice and two extremes. we should have the one virtue (i.e., p.eyaAO'TTpl'TT€La 
and only the one of the two vices which was the more akin (i.e .• the vice of being {JavavCTos). 
The notion moreover that the doctrine of related vices is derived from the doctrine of the 
mean gains some support from the fact that the doctrine of related vices occurs in its most 



JAMES A. COULTER 203 

Let us consider now the meaning of the terms f.LEtpaKunoE:; and 
t/JvXpoT7J:;. No small confusion has resulted from Philipps on's analysis 
of the categories of stylistic vices which are found in rrEp~ "YtP0vs 
3.3-4.15 If Philippson is correct, an important consequence is that we 
should be compelled to view f.J.EtpCtKtWOES and tPvXpOTT}S as two entirely 
distinct vices. Yet, as Mutschmann has shown, this is unlikely. For this 
scholar argued16 that in the discussion of'T6 t/JvXpoy17 which occupies 
chapter 4, this stylistic quality is characterized by two terms which 
are, for different reasons, closely connected with the supposedly dis­
tinct f.LEtpCtKtw8ES: the one ('T6 7I'Ctt8Ctptw8'U'TCt'TOV 4.2), because it is 
synonymous; the other (f.LtKpot/JvxLav 4.7), because it had earlier [3.4] 
been listed by Longinus as an attribute of 'TO f.J.€tpaKtw8€s. Conse­
quently-and Mutschmann's point is well taken-if in an account of 
'T6 t/Jvxpov these terms can be employed with perfect appropriateness, 
it would seem wrong to attempt rigidly to distinguish this latter term 
and 'T6 f.LEtpaKtw8€s. 

Moreover, and this Mutschmann did not note, the two terms ,vere 
used by the rhetoricians apparently as synonyms. Dionysius, for 
example, in a discussion of a strained use of stylistic figures, speaks of 
'Ta f.J.EtpCtKu;'07J 7I'aptaa Kat 'Ta t/Jvxpa aV'TLeE'TCt, [De Dem. 20 (U-R p. 171)]; 
surely, there exists no strict distinction between 7I'dptaa and aV'TLeE'Ta 
which would require terms of clearly separate significance. In just the 
same way, Quintilian [4.1.77] condemns Jrigida et puerilis ... ajJectatio. 
As to ajJectatio, moreover, there is evidence that we should consider 
as related in meaning to t/JvXpoT7Js and 'TO f.LEtpCtKtWO€S the term 
KaK6~7JAoV [3.4]; at 8.3.56 Quintilian notes: KaK6~7JAoV, id est mala 
affectatio. This is, in any case, a conclusion to which we should have 
been led on other grounds, since KCtKO'1'}AO:; was Demetrius' designa­
tion Echo 186] for the aberration of the stylistic 'character' (viZ., 
xapaKrryp y'Aacpvp6:;) which is the most closely akin to, if not identical 

developed form in Demetrius, who without question represents a rhetorical tradition 
which is overwhelmingly Peripatetic in character. 

16 "Zu IIepi "YI/lovs," RhM 74 (1925) 267-279. Philippson's thesis is further supported by 
Grube, AJP 78 (1957) 362-364. Munno essays a refutation (Mondo Classico 2 [1932] 309-312). 

18 Op.cit. (above, n.7) 88-91. 
17 It should be noted that in Aristotle's Rhetoric the term TO I/lvxp6v has a quite difIerent 

sense, and is in fact very close to what Longinus means by 'tumidity: In Rhet. 3.1405b34-
1406b19, the term designates, among other things, a tasteless use of metaphor. Theophrastus 
(cf. Demetrius ch. 114) appears to have used the term in a sense closer to that of Longinus. 



204 llEP! ygFOY£ AND ARISTOTLE'S MEAN 

with, Longinus' ~OV.18 And it is, we should remember, above all in the 
pursuit of T() ~ov that we lapse into if;vXp0Trjs. 

Furthermore, if TO ftELPUKLWOES is not identical with TO if;vxpov. 
where in the treatise was this important topic discussed? For it is im­
portant, as is shown by the statement TO OE ftEtPUKLWOES cXVTLKPVS 
lJ1TEvuVTtov TOLS ftEyH)E(1L 3.4. The problem of the lacunae has no rele­
vance here, for TO ftELPUKLWO€S is a topic pertinent to Longinus' 
discussion of vices in chapters 3-5 and it is difficult to imagine its 
having been introduced elsewhere. The portion of the treatise which 
preceded was devoted to a characterization of <tumidity,' and this 
subject is clearly dismissed as Longinus moves on to <puerility: 
There is no indication that the matter has already been discussed, 
since Longinus, apart from showing us how this aberration arises, is 
also at pains to give us a definition of <puerility: This is surely his 
intention in 3.4 (Tt 7TOT' ovv KT'\.), and it would be a serious error here 
to separate if;vXp0Trjs and TO ftELPUKLWOES, since <puerility' appears to 
be Longinus' term for the whole childish process whereby an author, 
in an overly clever and academic spirit, aims at a novel effect but 
only manages to err disastrously. Clearly, if;vxp0Trjs denotes the end 
result of the process for the whole of which Longinus employs the 
term ftELPUKLWO€S, and it is difficult to see how we can legitimately 
separate these two terms. 

It would therefore seem mistaken, in the light of this evidence, to 
construct, as Philippson did, a dubious theory of two ylll1], each con­
sisting in turn of three E£07J. 8uTlpov of 4.1 is clearly appropriate and 
cannot be used as evidence for two ylll1], since 7Tuplv8vpaov is intrusive 
and considered only briefly. And not, of course, because it was un­
important, but because, connected with the topic of TT(x8os, it was too 
important to deserve less than special treatment. It is also difficult to 
see how, if TO if;vxpov were a second and distinct ylvos, Longinus could 
say [3.5] TOVTCp 7TupaKELTuL TptTOV T£ KUKtUS El8os. TOVT<p surely refers 
to the TOVTO TO ylvos of the preceding sentence, and whereas I have 
argued that TOVTO TO ylvos assumes both TO if;vxpov and TO ftELPUKLWOES, 
Philipps on contended that it designated TO if;vxpov only. If, then, 
TO if;vXp6v is a different ylvos, in what sense can 7Taplv8vpaov, an eloos 
according to Philippson of the first, unnamed ylvos, be said to be 
<allied' (7TupaKELTuL) to <frigidity'? Finally, the distinction between 

18 Demetrius ch. 128; this fact is noted by Russell (op. cit., above, n.7) ad 3.4 TQ pUJ1TtKQV 

KCX~ KCXKO'''7Aov. 



JAMES A. COULTER 205 

YEVOS and EtSoS, upon which Philippson builds, is perhaps too un­
certain to allow us to form a theory on its basis,19 a theory which, in 
any event, is most improbable in view of the evidence we have been 
considering. 

To summarize. It appears certain that in 3.3-4 Longinus is employ­
ing techniques of literary analysis, and categories to which these give 
rise, which were derived from Peripatetic speculation on style, or, 
more precisely, from chapters 2 and 12 of Book 3 of the Rhetoric, and 
from later developments in the Peripatetic tradition, such as is 
evidenced in Demetrius and Horace. There is moreover evidence, 
afforded by the diction of the passage, which suggests-and this point 
will be taken up shortly-that Mutschmann was correct when he 
argued that Longinus knew the doctrine of the mean in its original 
setting, i.e., Books 2 and 4 of the Ethics. 

Be that as it may, the specific terms of Mutschmann's analysis can­
not be accepted as valid. For, as we saw, the vices whose genesis 
Longinus considers fall into two distinct groups: the one consisting of 
T6 ol8ovv and its opposite aaBtv€L(x or g'Y]pbT'Y]S, both of which are akin 
to <sublimity' CfLEydJo') 3.3 or TU vifrrJ 3.4); the other, viZ" TO fLELpaKLWOE') 

or ¢roxp6T'Y}s, aberrant products of a search for 'pleasantness' (T6 ?jov 
3.4). Consequently, Mutschmann's assertion that Longinus conceived 
of his virtue of sublimity as a mean which is flanked on the one side 
by T6 ol8ovv as an excess, and, on the other, by TO fLELpaKLWOES as a 
deficiency. cannot be strictly correct. 

Yet Mutschmann's evidence, even though it was adduced in support 
of an analysis which was in itself inaccurate, should not for that 
reason be ignored. For his analysis was inaccurate only in its precise 
terms; the general conclusions which it suggested, i.e., Longinus' 
familiarity with Peripatetic theory, was, as I hope I have shown, 
entirely justified. Mutschmann's evidence, moreover, was important 
not only because it pointed quite strongly to a Peripatetic context, but 
also because it suggested that the semblance of the mean was being 

19 These terms are in one instance used interchangeably in Aristotle; see Bonitz, Index 
Arist. 151b47-49. It is also not unlikely that Longinus used Y€Vo, in the sense of El80, under 
the influence of the Latin, as genus was a possible rendering of the Greek €lao, when this 
meant "kind"; cf Cicero Tim. 21 tertium materiae genus with Plato Tim. 35a3-4 TplTOV ••• 
oValar; £100,. For possible influence of Latin usage on Longinus, see W. Rhys Roberts, 
Longinus on the Sublime2 (Cambridge 1907) 188; H. J. Edmiston, "An Unnoticed Latinism in 
Longinus," CR (1900) 224; and J. C. Kamerbeek, "'EmKEl,..Eva (IIEpL ·y.povr; xxxiv.2)", 
Mnemcsyne 12 (1959) 128. My own guess is that Longinus wished to avoid using the term 
£100, twice in such close succession. 



206 llEPI Y210YE AND ARISTOTLE'S MEAN 

evoked in the discussion of <sublimity' and its attendant vices at 3.4 
(aMa 'TO p.~v ol8ovv K'TA.), even though there can be no question here of 
a derivation of the mean along strict Aristotelian lines, as was 
argued by Mutschmann. 

Now the suggestion that the semblance of the mean was being 
evoked at 3.4 must, of course, presuppose that the discussion of <sub­
limity' and its attendant vices has intentionally been cast by Longinus 
into a form analyzable from two standpoints, the one of which rests 
on the distinction between the <pleasant' and the <sublime' examined 
above, the other of which is reflected in Mutschmann's mean and is 
of a composite and artificial nature, since its extremes would be derived 
from two distinct categories, viZ" the <pleasant' and the <sublime: A 
discussion of this composite mean, and a consideration of the motives 
which led Longinus to employ it, will occupy the second section of 
this paper. Let us for the present examine some further items of 
evidence, some observed by Mutschmann, some not, which lend 
support to his thesis that Longinus knew the Ethics at first hand. 

With reference to the statement in chapter 5 (acp' cEv yap ~p.'iv 'Taya8a 

K'TA.), where Longinus advances the notion that both virtue and vice, 
success and failure, derive from the same source, Mutschmann com­
pared Ethics 2. 11 03 b6-8 : E'TL £K 'TWV athwv Kat 8La 'TWV a~'Twv Kat yiVE'TaL 

~ , , '.J.O' ., '" , " M h . 7TaO'a apETTJ KaL'f' ELpE'TaL, OP.OLW~ OE KaL 'TEXV7J. utsc mann qUIte prop-
erly argued that it was not likely that the similarity of concept could 
be fortuitous. Again, in the passage which was his chief item of evi­
dence (aMa 'TO fL~v ol8ovv K'TA. 3.4), Mutschmann argued that 'TO 

p.ELpaKLw8ES was in the strict Aristotelian sense the extremity the 
more opposed to the mean (aV'TLKpVS (J1TEVaVTlov 3.4). The notion of the 
one extreme being the more opposed to the mean is, of course, an 
important feature of Aristotle's discussion of Book 2 of the Ethics; cf 
2.110sb35 to 1l09a2: 7TPOS 8~ 'TO p.lO'ov aV'TlKEL'TaL p.aMov £cp' cEv p.~v ~ 
;MELI/lLS £cp' cEv 8~ ~ {mEpfJoAf]. 7TaplvOvpO'ov, as I have noted above, 
Mutschmann considers an intruder in the scheme. 

There are further items of evidence, not noted by Mutschmann, 
which lend support to his contention of Aristotelian influence. In the 
summing up [5], which follows the section on the vices akin to <sub­
limity', Longinus says: 

'" , , ~,,~ '" ~ \. '0 8 "'. <I OLOTTEP avayKaLov "JV'I OLaTTopELV KaL V7TO'TL EO' aL, O£ O'TOV 
, \, , ,~ •• 1_ \ ~ '.J. ' 'Tp07TOV 'Tas avaKEKpap.Evas KCXKtCXS 'TOtS V'!"II\OtS EK'f'EVYELV 

8vvap.E8a. 



JAMES A. COULTER 207 

This parallels, surely, the statement earlier in the same chapter 
examined by Mutschmann (i.e., &t/>' ~V yap KT:\.), and is noteworthy 
because of its Peripatetic phraseology.20 

Moreover, the assumption of Peripatetic context would make it 
clear, for example, why the one fault (TO oloovv) arises from a desire 
to Houtdo the sublime." For why is Gorgias censured for 'tumidity' of 
style, when he calls vultures "living tombs" (EfLtfvXOL T&t/>O£ 3.2)? 
Metaphor is, after all, a feature of the 'sublime' manner which 
Longinus seeks to define. Ought it not to follow that a richly meta­
phorical style deserves all the more to be called 'sublime'? The answer 
is, of course, No, and Longinus gives a clear answer in terms of the 
concept of propriety. If, Longinus says, excessive and tactless use of 
metaphor is ill-suited to poetry, which by nature admits more UTofLt/>o~ 
(3.1), it can scarcely be appropriate (apfLOU€L€V 3.1) to prose. The con­
cept of propriety as a criterion by which a given quality, essentially 
neutral, acquires its value, and a sense for the appropriate moment 
which is allied to this concept, are in this context clearly reminiscent 
of Aristotle's discussions in the Ethics and Rhetoric.21 For when our 
sense of propriety fails, those very qualities which would constitute 
virtue become the ingredients of vice (cf Ethics Z.1109aZOff). 

Finally, in the passage [15.3] where Euripides, a poet very little in­
clined by nature to 'sublimity,' is praised for boldly essaying the grand 
manner, we are reminded of Aristotle's discussion of the second rule 
for hitting the mean [Ethics 2.1109b 1-5]: UK07TEtV DE DEt 7TpO~ eX KaL mhoL 

, ~A.. I' "" , ' ''''' , A.. ' ", €VKCt.T""'{JOpOL €UfL€V. Ct.IV\OL yap 7TpO~ alV\Ct. 7TE'f'VKafLEV . . . EL~ TovvaVTLOV 
(i.e. to the one to which we are prone) 8' EavTov~ &t/>€AK€W 8€~. In the 
same way, Euripides ifKLUT& y€ TOL fL€yCt.Aot/>vTj~ WV 0fLw~ T~V atho~ Ct.VTOV 
cpvu£v (c[ 7T€cpVKCt.fL€V) EV 7ToAAo~~ y€v€uOa£ TpaY£K~v 7TPOO7Jv&YKCt.a€V (c[ 
Jt/>€AK€£v) [15.3]. In both cases, success is achieved by bending one's 
nature in the direction to which it is the less inclined. 

20 With aVIXK€KplX/-dvlX, cf Poetics 22.145Sa31, O€r K€Kpau8lXl TrW,. The significance of (EK) 
</J€Vy€LV as a technical term has already been noted (above, p. 200). 

21 E.g .• Rhet. 3.1404b3--4 p.~'T€ 'TIXTrWI.qV p.~'T€ 117rEP 'TO &~lwp.lX. aAAa TrptTrOVUIXV. Also cf ibid. 
140Sb13 app.oTT€L with app.oumv 3.1. For 'sense for the appropriate moment,' cf 2.2 1'(lv E.p' 
Jxau'Tov KaLpov with Rhet. 3.140sbl-2 'TO OE dJKa{pws 71 p..q wKatpws xpfjuBaL KOLVOV aTraVTWV 
7'WV €lowv Eunv. In this context, Longinus' use of the word aKpOTT}S [1.3] to characterize 'sub­
limity' is perhaps noteworthy. Aristotle observes (Ethics 2.1107a6-S) that although Ka'Ta /LEV 
7'.qv ovulav Kat 'TOV AOYOV 'TOV 'TO 'Tl1}v E lVaL AE'yoVTa virtue is a mean, it is Ka'Ta • • . 'TO apLU'TOV 
Kat 7'0 d; an aKp0TT}'; cf. however, D. H. Dem. ch. 2 [U-R p. 130] for the use of this term 
where there is no question of a Peripatetic context. 



208 rrEPI YlJIOYE AND ARISTOTLE'S MEAN 

II 
If the preceding analysis is correct, there are two problems which 

require attention. First, what is the significance of the two categories, 
vif1o~ and Tb ~ov, which emerged as a result of our own analysis? The 
second problem, which concerns Mutschmann's apparent mean 
("apparent" because constructed out of elements belonging to two 
separate categories), I should like to defer for the moment. 

At the very outset it should be made clear that the appearance in 
chapter 3 of ;;if1o~ and TO ~S6 as two distinct, yet allied stylistic cate­
gories is not fortuitous. Several parallels, in fact, occur within the 
IIEpt "Yif1ov~, as well as in the work of the ancient literary critics, 
particularly Dionysius of Halicarnassus. Let us consider the following 
passages from the IIEpt "Yif1ov~; it should be noted that the fL€{'a>..7]yopla 

of 39.1 (if Toll's conjecture is correct) and Tb ()avfLamov of 1.4 are 
merely synonyms of ;;if1o~ or v~/...6v, as is XcX.Pt~ (1.4) of TO ~ov: 

"() "-I,.' , , () , '\ \ o EV E7TL'j'OpOV €L~ uvvTaYfLaTWV KaTOp WULV Ta TE KalV\7] 
~. , \\ fI.l. \ \ , C c t' I [5 1] T7]S EPfL7]v€La~ KaL Ta v.,.1] KaL 7TpO~ TOVTOL~ (n 1]00"'«, • 

() ,.. , H .... '\ \ f' ,.... 8 
aVfLa fL EXEL, ••• 7TW~ KaTa TOV 7]fLETEpOV aLwva 7n «-

, '\ "" \ \ ,~ .... , " .... va, fLEV E7T aKpov KaL 7TOI\L'TLKaL, opLfLHaL TE Ka, EV'TPEXEt~, 
\ ''\ \ c ~ , \' ".L -(.t. '\ '''' \ \' Kat p.a'''ClTa ."pOS 1]OOJlaS I\oywv €v.,.opoa, v.,.1]"a, OE I\Lav 
\ • 'e \'" I " -KaL VTTEpfL€{'E €I.~, 7T1\7]V Et fL7] TL UTTavLOV, OVKETL YEVVWV'Ta, 

cpvaEL~. [44.1] 
.,,&80s oE iJt{lovs fLET€XEL ToaovTov, o7T6aov ~80s ~80Jlfls. 

[29.2] 
" ,\ 8'" "~ .... f ~ I"L •.• ov fLOVOV EU'TL tJ'n OVS' Kat 1]OOV'1S' 7] apfLovLa 'j'VUL-

, '()' , \ \ \ \ '\ , \ 18 KOV av pW7TOL~, alV\a Kat p.€yal\1}yoptaS' Kat tJ'a. OVS 

()avfLaaTOV'TL opyavov' [39.1] 
7TcX.VT7J O€ yE uvv 11CtJ'/....qfn TOU tJ',8aJloO Kat TOO tJ'pOS 
, '\ ~, 8' [1 4] xapw aEL KpaTH TO avp.«CI'OJl,... • 

It is soon apparent, moreover, that with this pair of stylistic qualities, 
i.e., with ;;if1o~ and Tb ~Sv, there are associated additional elements. 
From the passages above we see that this contrasting pair is in itself 
incomplete; the full scheme is vif1o~ - 7TcX.8o~ - £K7T/...7]g,S', as 
opposed to ~ov - ~8o~ - 7T€LfJJJ. Indeed, the former part of this scheme 
appears explicitly in Longinus' description of De most henes' lightning­
like eloquence [12.5]: 

\ '" \ ~ LI fJ ~ \ ".1. \ • , KaLpo~ OE TOV 7]fLoa EVLKOV fLEV v't'0v~ KaL V7TEpTETafLEvov 

£V TE TaL~ SELvwaEa, Kat TOL~ acPoSpoi~ 7TCxfJEUL, Kat £v()a SEt 
\' \ \ '\ , \ ~l: 

TOV aKpoa'T7]V TO UVVOI\OV EK7TI\7]<:.aL. 



JAMES A. caUL TER 209 

It can be shown, moreover, that Longinus is employing a complex 
familiar in ancient literary criticism. The triad vif;os - 7TaeOS - ;'K7T}..7]g,S 
is, in fact, already present in Aristotle, at least by implication;22 while 
the entire scheme of two contrasting triads is employed by Dionysius 
in the De Lysia 13 [U-R p. 23]: 

,~, r/ , "'" "8 I, 8' rl , .... ovo , Wa7T€p €V 7"Ot~ YJ €atv €a7"L TTL aVYj, ov7"w~ €V 7"OL~ 

'e ' "", < < I;' ~ ,~ , I e 7Ta €aLV Laxvpa ovo w~ YJovvaL Kat 7TELaa£ KaL xapL€Vnaaa aL 
1;'1 "f3' (J" I OVVa7"aL, OV7"W Laaeta aL 7"€ KaL 7TpoaavaYKaaaL. 

Dionysius' laxvpa alludes to his stylistic category, taxv~ or o€Lv67"YJ~; 
this is close to Longinus' vif;os, when vif;os contains an element of 
strong passion (cf. below, n.28). As to f3LaaaaeaL, II€p~ "Yljlovs 1.4 makes 
it clear that f3{a (cf. f3tav eXfJ-axov) is a constituent element of ;'K7T}..7]gLS. 

It is clear from this evidence that the contrast between the 
<pleasant' and the <magnificent' in style, which had been previously 
unnoticed in 3.3-4, was a familiar one, and one which ultimately, I 
suggest, derives from Theophrasrus.23 

But to return to Mutschmann. That vif;o~ should in his analysis be 
flanked by two aberrant stylistic qualities, the one of which (7"6 oloovv) 

is allied to fJ-Ey€8os, the other (7"6 fJ-€LpaKLWD€S) to 7"6 ~ov, now seems 
in itself less curious. Moreover, since vif;os (-7]) and fJ-Ey€8os (-7]) are for 
Longinus synonymous, it is not strange that he should bring 7"6 olooiJv, 

which is an aberrant product of a search for fJ-l.y€eOS, into relation with 
TeX v!f7J as one of its vicious extremes. But why 7"6 fJ-€LpaKLWS€s (and TO 
ifroXp6v)? For the discussion of 7"0 if;vXp6v which follows [4] seems to be 
only in terms of 7"6 ~ov,24 and while it is a concept associated with 

22 See Rhet. 3.1408bll-12, 'Ta 8~ Qv6fLa'Ta ••• 'Ta gEva (for these as elements of the 'sub­
lime' style, cf. ITEP' "Yopov, 30) fLc7.A£aTa apfL6'T'T€£ Myov'T£ 7ra8'Y}'T£KW" and ibid. 1408a23-25 
(]VVOfLO£07ra8E'i 0 aKovwv &.-1. 1'0 7ra(J'Y}'T£KW, Myov'T£ •.• aU) 71'0'\'\0' Ka'Ta7TA-q'T'TOVa£ 'TOVS aKpOa'Tas 

80pv{JovV'T€S. 
23 It was argued by Stroux (De Theophrasti Virtutiblls Dicendi [Leipzig 1911] ch. 2) that 

Theophrastus in his ITEP' AEgEW, subdivided his fourth virtue of style, viZ., KaTaaKEV~ or 
K6afLos, into ~8V and fLEyaA07rP€7TES. The first three Theophrastean virtues, according to 
Stroux, were aacP-qv€ta, JAAlJv£ap.6" and 7TP€7TOV. What stylistic qualities were indicated by 
fLeYaA07rP€7rfS it is not easy to reconstruct with precision. By a process of elimination, how­
ever, i.e., by eliminating the first three virtues and ~a,;, it appears that fLeya.\07TP€7T€S 

designated qualities very much like those for which Longinus employed the term vt/11J),6v. 
And we should also remember that fLEyaA07rpE7T€S is itself a synonym for v.p'Y}AOv; cf. 12.3 
and 30.1; also D. H. De Lysia ch. 13 [U-R p. 23] vop'Y}A~ 8~ Ka, p.eyaAo7TpE7r~S .•• For a recent 
discussion of Theophrastus' ITEP' ..1€gEWS see G. Kennedy, The Art of Persuasion in Greece 
(Princeton 1963) 273-284 and the scholarly literature cited there. 

II' The chief cause of opvxp6'T'Y]'3 according to Longinus is a search for new and clever ideas, 
and his examples show that 'frigidity' occurs chiefly when an attempt at wit misfires. 



210 nEPI YP'OYL' AND ARISTOTLE'S MEAN 

vifios, it is also, as the evidence cited above demonstrates, quite dis­
tinct. The conclusion suggests itself that Longinus, employing ter­
minology which reflected two well recognized areas of style, sought 
to imply a stylistic category which was constituted of both, and that 
in order to do so he had recourse to the form of the Aristotelian mean. 
Why this should be so we shall now consider. 

III 
An answer to the problems raised by the preceding analysis im­

mediately suggests itself if we recall the character of the treatise. It is 
a polemic aimed at discrediting Caecilius' adverse criticism of Plato's 
style, and at pointing out the absurdity of Caecilius' omission of 71'aOos 
in his discussion of 'sublimity'. 

Caecilius, as Longinus tells us [32.8], considered Lysias an author 
without fault, and for that reason superior to Plato, who was, as 
Caecilius maintains, guilty of numerous stylistic lapses. Longinus 
replies that even if Caecilius' charges were true, which they are not, 
plato would be superior to Lysias merely on the grounds that he 
possessed vifios, a virtue which redeems all stylistic faults [36.2; cf 1.3]. 

This entire discussion [32.8-36] presupposes the elaborate system of 
stylistic virtues which had been worked out by the time of the 
Augustan period, although there are hints of it already in Cicero. It is 
also familiar from the critical works of Caecilius' contemporary and 
friend, Dionysius.2s We are not, to be sure, equally well-informed 
about Caecilius' system of virtues and vices, if indeed there was any. It 
is clear, however, that Caecilius considered vif;os a virtue (as did 
Dionysius). This we can infer from Longinus' observation [1] that 
Caecilius failed to show the reader the means to acquire vif;os; rather, 
Caecilius merely illustrated what it was by means of numerous ex­
amples. Needless to say, Longinus' charge would be senseless if 
Caecilius had condemned vif;os as an undesirable quality of style. For 
Caecilius, vif;os was therefore a virtue of style, although in no wayan 
indispensable one, since Lysias, his exemplar of perfection, was with­
out it. That Caecilius explicitly maintained that Lysias lacked vif;os is 
not known, since the fragments ofCaecilius' two studies of Lysias have 
nothing to say on this particular matter. Dionysius, however, denied 
this range of style to Lysias, and common sense requires that if vif;os. 

25 See s. F. Bonner, The Literary Treatises of Dionysius of Halicarnassus (Cambridge 1939) 
18-24. 



JAMES A. COULTER 211 

ILEyaAo77pE7TELa, etc. had any stable meaning, they could not have been 
used by Caecilius to characterize Lysias' style. This was ~Sv, not V!f'T}AOV 

or fLEyaAo7TpE7TES; it possessed xapLS, not v!fos. 
Lysias' charm and grace are the burden of Dionysius' praise (De 

Lysia 10 [U-R p. 18])-a view which Longinus, too, shared, as is clear 
from a remark in his characterization of Hyperides [34.2]: "He posses­
ses all of Lysias' virtues and his graces (xapt'Tas)." In other words, 
Lysias, an author rich in 'charm' and 'grace,' but devoid of 'splendor' 
and 'sublimity: is accorded by Caecilius a place in the critic's hierarchy 
superior to that of plato. The reason that Longinus found this con­
tention insupportable will be obvious to every reader of the treatise. 

What I do suggest, however, is that Longinus' dissatisfaction, and 
the polemical interchange which it engendered, found expression also 
in the precise details of rhetorical theory reflected in the passages 
which we considered earlier. Furthermore, Longinus appears in chap­
ter 3 to be drawing upon a tradition, which can be traced back at least 
as far as Cicero, according to which distinction in style-anything, that 
is, which remains after the bare requirements of communication have 
been discharged-was divided into the pleasant and the magnificent.26 

Although some occasions might demand more of the one than of the 
other, both were, and should be, constant features of a style which 
pretended to more than mere clarity and correctness: the best orator 
would have both at his command. 

Now the passages quoted above show that Longinus viewed the 
matter in somewhat this light: Kat Ta vif;TJ Kat 7TPOS TOllTOtS al ~oovcd 
[5 1] d " " r \:' ~ r r '-I.. "e' " , , . ,an ov fLOVOV EaTL . • • 'T}ooV'T}s 'T) apfLovta <rvatKOV av PW7TotS, alV\a 

Kat fLEyaA'T}yoptas ••• opyavov [39.1]. Moreover, despite his views on 
the importance of sublimity, the admiration of Longinus for a charm­
ing and graceful style, as is shown by his characterization of Hyper ides, 
is quite sincere [34.2], as is his recognition of the real way Demosthenes 
blunders when he attempts to be jocular and pleasant: "he does not 
excite laughter but rather becomes the subject of it" (ou YEAwTa KtJJEL 

fL&)J\OV ~ KaTaYEA&Tat 34.3). Both qualities, therefore, should ideally 
be present in style; it is an acknowledged defect of Demosthenes that 
he lacks a graceful wit. However,-and here Longinus parts company 
with Caecilius-if one were compelled to choose one of the two, and to 

26 See Orator 79 and Kroll ad loc.; cf. also Part. Drat. 6.19-22. See also above n.23, where 
Stroux's conclusions suggest that very much the same idea was already present in Theo­
phrastus' IIEp1 A,f~EW<;. 

G.R.B.S·-5 



212 lIE PI Y1JIOYE AND ARISTOTLE'S MEAN 

say that this is the expendable quality, even without this an author can 
be supreme, then one must surely eliminate 'charm,' not 'sublimity: 
This is one obvious source of Longinus' displeasure with Caecilius for 
his elevation of Lysias to the level of perfection and his consequent 
preference of the orator to plato. Lysias, according to Caecilius, ex­
celled Plato; and yet Lysias did not possess v,pos, whereas Plato did. 
This surely was not a critical position which Longinus could calmly 
and decently endure. It should be noted, moreover, that this same 
outlook characterizes the aVYKpuns of Hyperides and Demosthenes 
[34]. 

It is, I submit, precisely into the framework of this literary debate 
that the section ought to be placed, to the Peripatetic character of 
which Mutschmann was the first to call attention. Its purpose was to 

enunciate the two terms of the debate, i.e., the 'pleasant' and the 
'sublime,' and at the same time to suggest, with the help of what I 
have called an «apparent" mean, the traditional notion that distinc­
tion in style was a composite of these two qualities, even though for 
Longinus the presence of 'sublimity' always redeemed any lapse or 
omission in the domain of the 'pleasant.' My own conjecture is that 
the authority of Theophrastus is consciously being invoked in this 
particular matter. This cannot of course be proven. Cicero, however, 
provides an interesting paralle1.27 

To come now to the second point in Longinus' critique of Caecilius' 
treatise, i.e., Caecilius' omission of 7TCf8os in his discussion of sublimity. 
There can have been two reasons for this, Longinus argues, and both 
are wrong. The second explanation put forward is that Caecilius 
thought that 71'a8os had no part at all in the virtue of sublimity (inci­
dentally, if he did, he would only be sharing Dionysius' opinion).28 
This, I suspect, was the interpretation to which Longinus inclined; at 
any rate, he devoted much more space to this alternative. Now we 
have already seen that for Longinus 7TCx8os and v,pos are complemen-

27 See Orator 79, where Cicero uses the authority of Theophrastus to support his conten­
tion that the Atticists' conception of style was inadequate. 

28 For Dionysius, l7cf80. was not an ingredient of iJ.po.; cf, for example, De Isoc. ch. 3 
[U-R p. 59], where the iJ.po. of Isocrates, who is an exemplar of this virtue, is praised, and 
the absence of TO l7a8",/TtKOV noted (ibid. ch. 2 [U-R p. 57]). Rather, it was associated with 
what Dionysius called luxU' or a"tVoT"'/'; cf De Thucyd. ch. 53 [U-R p. 412] T1}v J~eydpovua.v 
T<X .".cf8"'/ aEtV0Tl}Ta. The iJ.po<; of Longinus appears to be a composite of both categories of 
Dionysius; in 8.2 it corresponds to the iJ.po. of Dionysius, in 8.4 to his luxu •. As Longinus 
states in his discussion in ch. 8, there are some kinds of iJ.po. which have .".cf8o., and some 
which do not. See L. Voir, Ll"tvoTl}', ein antiker Stilbegriff (Leipzig 1934) 47-53. 



JAMES A. COULTER 213 

tary concepts, and that the two imply the further notion of EK7T>""f'J~L~. 
Longinus allows, to be sure, that there are examples of the sublime 
which have absolutely no trace of passion. Yet he does insist [8.3] that 
nothing contributes so effectively to 'sublimity' as noble and exalted 
passion. There are several reasons, but the fundamental one surely is 
that 7TlxOOS is one part of us in which the divine manifests itself. From 
this fact derives the poignancy of the last chapter, where Longinus 
points out that it is not political oppression which has caused the 
dearth of great literature, but the corruption of our passions by a pros­
perous material environment. For in a world where our passions, 
sullied and cut off from their divine source of nourishment, have re­
placed what is truly god-like with the spurious divinity of wealth, 
there is no possibility of a renascence of literary greatness. 

It is in this light that we should reconsider the beginning of chapter 
44, where the contrast between the 'pleasant' and the 'sublime' is 
most clearly expressed. In the context of the analysis of the causes of 
cultural decline which immediately follows, Longinus' polemic 
against Caecilius concerning the relative merits of 'charm' and 'sub­
limity' assumes a quite different, a more urgent and profound charac­
ter. For to exalt, as Caecilius had done, the 'pleasant' over the 
'sublime,' and to maintain further that pathos had nothing to do with 
this latter quality, was not merely a matter for a fellow critic's cen­
sure; it was also to acquiesce in and encourage the root evil of the time. 
Caecilius, by saying that 'sublimity,' with its all but invariable com­
ponent of transcendent passion, was not necessary for literary great­
ness, surely seemed to Longinus to be denying the potential function 
ofliterature to cleanse men of their avarice and self-gratification. 

In conclusion, it need scarcely be emphasized that Longinus was no 
ordinary rhetorician. The way in which he adapted to his own extra­
ordinary requirements the often arid and doctrinaire details of 
rhetorical and stylistic theory shows this quite clearly. Yet we must 
remember that Longinus was, first of all, a rhetorician, and if this had 
not been the case, if he had not had at his ready disposal the elements 
of a long and fully elaborated tradition, his conception of the 'sublime' 
would not have been clad in precisely the garb in which it was, nor 
without the Peripatetics would his polemic with Caecilius have 
assumed the particular form that it did. 

COLUMBIA UNIVERSITY 

August, 1964