You are on page 1of 4

The Arabic Version of Galen's De Sectis ad eos qui introducuntur

Author(s): J. S. Wilkie and G. E. R. Lloyd


Source: The Journal of Hellenic Studies, Vol. 98 (1978), pp. 167-169
Published by: The Society for the Promotion of Hellenic Studies
Stable URL: http://www.jstor.org/stable/630206
Accessed: 20-09-2016 16:10 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Cambridge University Press, The Society for the Promotion of Hellenic Studies are
collaborating with JSTOR to digitize, preserve and extend access to The Journal of Hellenic Studies

This content downloaded from 158.251.134.79 on Tue, 20 Sep 2016 16:10:01 UTC
All use subject to http://about.jstor.org/terms

NOTES
and
Days.
different

The
idea
of
type
by Dr Lyons);
or by the consultation
of further
Arabic
animals
is
set
manuscripts, which showed
faults of
transcriptionout
in the

Semonides
women

are

167

who

7.
Phocylides'
manuscript
from which our first transcript was made. To f

explain what we from


have found with respect
to Hunain's
come
the
7r

describedchoice
in
of individual
similar
words, we must describeterm
the method

sow-

and

bitch-women,
how
we adopted. One of us (J.S.W.) made a translation,
as

nearly as possible literal,


from the
Arabic into English,
correspond
rather
to
Semonid
scrupulously Er'
avoiding any aOv
consultation of the
Greek text.
ydp
KaKOV
o07
oViv
respectively. WePhocylides
hard
then met and compared this English version
with the
allows
himself,
in again
giving
his
Greek. It happened
and again that an English
word
porcine
feature.
Whether
was found to represent
the Greek word tolerably well,
Semonides
or
a
commonp
but noton
exactly. Reference
to the lexicons showed, howcannot
be
decided
with
certai
ever, that the English
word chosen represented
merely
To
sum
up:
the
evidence
is
one of the possible
meanings of the Arabic, and
amongco
esis
that
the the
hexameter
frag
others listed was to be found the exact equivalent
of
the Greek.
come
from
a
gnomic
poem,

first
half
of
the
sixth
century
A persistent
source of scepticism
as to the value of
represented
as
giving
t
Arabic manuscripts
of translations advice
from the Greek is the

of
topics
including
belief that, however good the how
original translationto
may
how
to
conduct
oneself
have been, the extant
manuscripts mugreatly in
infected a
political
tensions;
the
differ
with glosses, introduced
in the course
of transmission,
poem

less

fits

well

archaic

well
when
seen
in
that no reliance
can be placed upon
the Arabic, as we
now

when
against
have it, as a seen
representation of the
original text of the

translation.
Greek
and
M. L. WEST

Bedford College, London

earlier

ori

We can say categorically that this is totally false of the

extant manuscripts of the treatise here considered; and


equally false of the manuscripts of the Ars parva with
which we are now engaged (It is also false of the MSS of
the De usu pulsuum and of the An in arteriis natura sanguis

13 Bergk appends to his collection of fragments a number of anony-

contineatur,
J.S.W.).
mous gnomic hexameters which may come from Phocylides.
Their
We have found perhaps five or six possible glosses, each
consisting of only a few words. Of these it is possible that

inclusion would not significantly affect the picture I have drawn.

one or two represent words present in Galen's Greek,


which have dropped out of the Greek MSS. It is, moreover, extremely difficult to distinguish between what is to
The Arabic version of Galen's

'De Sectis ad eos qui introducuntur

be considered a gloss (where only a few words are concerned) and what is a legitimate explicative translation. In

one case, where we have conceded a possible gloss, Dr

Lyons considered that what we have is introduced simply


In this study' we have compared an Arabic translation
to avoid a gross inelegance, which would result in Arabic
with a well-edited Greek text, in the preparation
of
from the adoption of a too direct rendering of the Greek
which, moreover, the editors have found no evidence of
syntactical form (see below on H 15.24).

any major lacunae in their MSS, and which we may

We defer the identification of the Arabic MSS to the


suppose to correspond closely with Galen's original comend of this paper and pass at once to our results.
position. It was, consequently, from the first improbable

that we should make any very striking discoveries:


(Note on abbreviations. H: G. Helmreich's edition in Claunothing was to be expected at all comparable with the
dii Galeni Pergameni Scripta Minora iii (Leipzig 1893). K:
indubitable evidence for the inversion of a leaf in the
Kuihn's edition, Claudii Galeni Opera Omnia, i (Leipzig
codex from which are derived our Greek manuscripts
of A: J. S. Wilkie's translation of Arabic version. The
1821),

Galen's An in arteriis natura sanguis contineatur (Furley and

Greek MSS are referred to according to Helmreich's


Wilkie, CR xxii [1972] 164-7).
edition, viz. L=cod. Laurentianus LXXIV 5, Ll=eius
For anyone interested in the Arabic translations of
manus prima, L2= eiusdem manus altera; M= cod. Mos-

Galen's works there was, however, a compensatory


quensis 283; m=cod. Mosquensis 51; V=cod. Venetus

advantage in the possibility thus presented of assessingbibliothecae


the
Marcianae V 9; v=cod. Venetus eiusdem

quality of the Arabic version. The reputation ofbibliothecae


the
282.)
translator, Hunain ibn Ishiq, has, indeed, already been

long established; and we were prepared to find his transH 4.22 f. [?aEL]. K 69.10 ;aLv. Marquardt ;aToEW . A:
lation at the least respectable; it is safe to say, however,
'be well acquainted with the diversity of states of the

that we have found it uniformly excellent. In the three or

air, of waters...' K's /aeTLV is ungrammatical. The

four cases where a sentence seemed to have been misunArabic suggests the presence, in his Greek text, of a
derstood, the initial impression of error has been dissiword (e.g. caToEws or qfaEov) governing dipwov,
pated by a more careful reading of the Arabic (confirmed
68rowv.. .: a double genitive would have been clumsy
but might help to account for the corruption.
I We must express our warmest thanks to Dr M. C. Lyons of Pembroke
College Cambridge, who has been most generous in allowingHus5.3
to [fpptaKov]. K 69.16 stet. A: 'this drug'. The Arabic

consult him on many particular points. He takes, of course, no responsisupports retention of /apt/aKov.
bility for any inaccuracies that may remain.
H 7.12 ff. The Greek text suggests

This content downloaded from 158.251.134.79 on Tue, 20 Sep 2016 16:10:01 UTC
All use subject to http://about.jstor.org/terms

greater bleeding if the

168

NOTES

patient
is
used
to
The
item but has a reference
to xwpasit.
at H 12.19 rather
than
no
equivalent
the
addit
one to xdpavto
at H 12.16 (cf.
17.8).

in two of our three Arabic MSS: 'but this has now

stopped'.
H 7.18 The Arabic has 'the method of making out (or

H 12.23 f. trELp6pa.vral y' ol p,v "rV Ka7ard &atrav


vooradrO.wv. rWV yE of pv (Aldine edition), voorapd'rwV

adrrcivTwv (vV). K 80.2 f. 1rr.pdv-raL -vv yE o1 ILLv Ka'a

inferring) that remedy and discovering it', which pro-

sa-raay vooITdtLvaiW divtrcv. A: 'and some of them


claim that these groups embrace the diseases of which
the treatment is by regimen.' The Arabic clearly supports H.

vides some support for epecTowso (edd. and other Greek

MSS) against alp&eowsr (L1).


H 8.1 and 3 IliXPL trav-yS (omitted by M at H 8.1 and by

LI-added by L2--at H 8.3). A: 'frorfirst to last' in

H 13.9-12. The Arabic translates the sentence -rIjv ylp

prising about this passage is that Galen should say that a


bite from a mad dog is the same as other bites 'from first
to last': whereas this is true, rather, of bites from dogs

H 13.1o v,3v. K 80.12 (with Lv) v, though K's Latin

both passages. The Arabic supports edd. What is sur-

$,.ypov7j
v ... .rrTOos ErTrt1.rr,.AEyL4EVov that Marquardt
judged
spurious.
translation has 'quoniam non sola nunc'. A: 'when it is
in a state of this kind it is not simply one.' No clear

other than mad ones (cf. H. 18.24 ff.). We would


expect-and perhaps might conjecture-/m) before

evidence of a vt3v, though 'now' would have been


clumsy here in Arabic.

Av-r77rroS in H 8.3).
H 8.21 ff. r1Tl rals 7lAudaLS Kat 7"ai gpatg Kat 7atS xcpaQt

H 13.18 f. 7w yap /iaAAov EVOXAOV7L Kat 7V KLV VVOV

EtrmfpovTr, 0ro7' oEat 7w laXvpoTEp?O., evavrtobvcOat. L2

EKaaVou
. Eprl7LEwov
4) ages
OepaTEILa.
'the treatment ... in7rW
accordance
with the
and theA:
seasons
of

has T( K^ VVw,, LV eo eto 8 o rt rxvpdoTEpov. K

the year and each of the things we have mentioned.'


The Arabic omits KaZ 7ratS G xwLpa and takes 'each of the

things we have mentioned' as coordinate with 'seasons

of the
(i.e. perhaps
H 8.23
S&' year'
'repL7plas
(omitted&K.TaroL9).
by L). A: 'by experience'.
Supports edd.
H 9.19 f. ,ra 8' ~4tTrr?> Trv rd pav otatha. V8J o a. (om. V)

rd %v (om. L) LMV T& ic ivv/1% v. K 75.15 rdr S, rCP


p1v 7oTatrfa. A: 'and as for the statement made by the
people who concede to the Empirics that these things

81.3
f. o0Tro 8 s r-rtV, dWS 7-r lXvpo-rEp w. A: 'the
resistance to the disease that afflicts him more and is

more dangerous, and that is the more powerful of the


two diseases, is more important than the resistance of
the other.' The Arabic supports H (or LV) against L2.
H 15.23 f. There is no sign, in the Arabic, of an equivalent
for 1rpdTm7r H 15.23 or vvv H 15.24.

H 15.24 There is no equivalent, in the Greek, for the


sentence 'on the contrary, the dispute goes back to the
words and to the deeds' in the Arabic, but as Dr Lyons
suggests this may be a stylistic expansion for the sake of

may be discovered by it [experience] but with the

balance.

qualification that they object to it [experience] that it is

not limited and takes a long time. . .' The Arabic


provides some support for H (cf. t5r' at H 9.13).
H o0.6 To-T 7e aa-aroa -ri9 v /qVow. After 'knowledge of the
nature of the temperament of the body' one of the three
Arabic MSS adds (what is probably a gloss: the text is in

any case difficult and as Dr Lyons has suggested dittography may be suspected): '[both that] which is natural
to it; and that to which it is transformed, so that it needs
to be transformed when it reverts to that which is
natural to it.'

H 16.6 pdovov omitted by MmVv, square-bracketed by H,


read by K at 84.10o. No trace of this in A. (Contrast 'by
reason alone ' for po'vov at H 16.3 and 4).

H Supports
16.18 &pylas.
K 85.2 (with v) .vepyelas. A: 'repose'.
H.
H 17.8 The Arabic order is that of Vv and K

86. I--opa. .. -.cALKlaS. .. Xwdpag (Xwpla K).

H 17.23
6 &ppwaorcvw].
omitted
byby
LMmV,
elva~ o[EtlaL
&ppwaordv
deleted by E.vat
Marquardt,
read
K

86.14 f. A: 'in whatever country he happens to fall sick.'


H 11.2 [d87Aov oVtEr6 trrpdytparos odsvo'8s] (cf. H 10.25). K Gives some support for 6 &ppwoard-v.

H 18.1 [ErttoKo1rToatJ]. LMmV omit. K 86.17 stet. MSS


77.12 exists
stet. Of
the Greek
MSS,
L omits o.Sevo's.
'there
no proof
whatever
demonstrating
some- A:
and edd. have no verb in the sentence at H 18.2 &p' odt'y
thing hidden.' The Arabic may suggest a text with at Kat 7a7a Tcrata. . . A: 'Now what will they say

least a8'Aov otSevos, unless, indeed, the Arabic tradi- about the parts of the body? Will not the consideration
tion has added the words from the previous context, H of them be superfluous.. .?' The Arabic suggests an
10.25.

cf.
H. 21.1). whether
in theor
first
or in the second
original
.i.KO.trrat
E7LTKOTEEV
(orsenrtroKoWtraoa,

H 11.8 f. 8v 87 <rcv> rawoLwvopLtov AdYov ' .val ract

(awdltvovov
L1Mv).
K 77.17-78.1
5V 88TheOawdLve.vov
Etval
/aoc. A: 'the
inference
to perceptibles.'
Arabic supports H.
H 12.2 dvooAoylav (with L). dvootoAoylav m, d Pohoylav
Mv, dvaAoylav V. K 79.1 dywvoAoylav. A: 'controversy'-some support for H.

H 12.3
bav.v (with
Mm).translation
K 79.2 (with
MSS)A:
/atLEv-but
K's Latin
has other
'apparens'.

tence.

H 18.9 det. K 87.7 (with v) omits. A: 'constantly'-

ports H.
H 18.19 o-t 8'. K 87.15 odT'. A: 'then you will have gained

an empty triumph'-supports H.
H 19.3 There is no equivalent, in the Greek, for the
Arabic: 'so that he did not discriminate the part in
which the bite was', a possible-though not very like-

'when the thing has become clear and manifest and


how it is has become perceived'. Supports H.

H 12.16
oi;0'dpav o-rE. Xypav and H 12.18 f. [r&s Jpas Kat
rdS Xdpag Kat] 7-r E7 (square-bracketed material omitted by V). K 79.13 and 15 f.: no square brackets. The
Arabic omits obv-E Xpav in the first passage (H 12.16)

and has 'the consideration of countries and customs' in

ly-gloss. The evidence suggests that Hunain found

this phrase in his Greek MS.

H 19.6 There is no equivalent, in the Greek, for the


Arabic: 'whose property was to draw out poison and to
alleviate (or dry) it', though this may be a gloss.

H 19.22 f. [K7card rrV7a]. K 89.8 stet. A: 'and there is no


difference between the two diseases in any way whate-

the second. Like H, the Arabic avoids repeating any

ver'-some support for K.

This content downloaded from 158.251.134.79 on Tue, 20 Sep 2016 16:10:01 UTC
All use subject to http://about.jstor.org/terms

NOTES

169

Arabe 2860 in the Bibliothique Nationale, and


MS 1075 of
H
19.23
Trdv-ror4E
(with
L
Garrett Collection
of the Library ways
of the University of
rcv'rri.
A: the
'in
all
th
Princeton.2 As is frequently
the
case, each number (2860 (wit
H
20.4
t
Vpcv.
K
89.13
and 1075)mention
is appended, not to a MS of a single treatise, but oth
you,
not
to
to a small collection or set of treatises which have some-

H
20.
16
rrpoolEaoEt
7as
p
thing in common,
and which aretimes
collectively referred to
raS
xwpag.
A:
'the
of
in the cataloguesof
as 'a manuscript'.
in
the
Arabic,
Ka
Irs
X
The Parisian
set (2860) consists
of four
H
21.4
vavTr
ov.
Lof translations
omits.
contrary

works what
by Galen (or, strictly speaking,
of five; since a do'.
first
to
you

and a second 'to Glaukon'


distinguished),
transcribed
H
21.18
There
is areno
equiv
by the same
scribe in the
year 676 of did,
the Hijra (A.D. 1277). it
Arabic:
'and
if
he
Of this set,
the first is the translation
the De sectis.
which
looks
like
a of
gloss.
terested

in

The Princeton set (1o75) is saidof


in the catalogue
(Hitti et
disease
the

al.) to consist of translations


of nine of
Galen's works, of h
23.15
and
24.5
which one (Ars
parva) was transcribed in A.H. I 13in
8 (A.D.
K
has
trapd
b6owv
t
1726); and
the remaining K
eight, in A.H.
572 (A.D. I 176/7). w
93.14-with
L2;
94.4

23.6

f.,

In the catalogue,
the translation
of the De sectis is the sixth
translation
of
93.14
'sec

'natural evacuations' on all three occasions.

of the set of nine. There is, however, in Hitti's catalogue a

mistake of considerable importance for us; since what was


H 24.22 [7raX~~w Kal]. K 95.16 stet. H notes 'Kat ante

written in A.D. 1726 was not, as stated, 'the introduction


TaxEws add. LMm'. A: 'they promise to teach us this art

quickly in six months'-supports K.

[to the set of nine treatises] and The Book of the Minor Art',

but the De sectis followed by the Ars parva. In other


H 25.22 (with L2MmV) Oeparrlas. K 97.3 (with L')

words, the translation of the De sectis, which lacks both


Oewptas. A: 'each of them requires a particular treat-

title and colophon, was mistaken for an introductory


ment'-supports H.
essay composed by the scribe of A.D. 1726.
H 26.6 [raAatj] (omitted by Lm, also K 97.11). A: 'Hippocrates already epitomized all those things. . .'. ThisConsequently, our two microfilms provide us with
three transcripts of the Arabic of the De sectis; one of Paris,
gives some slight support for a temporal adverb,
and two of Princeton.
although the qad (here translated 'already') may have

been added, as is sometimes the case, merely forOf the three transcripts, those in Princeton are superior

to that in Paris, in so far as the Parisian transcript has


important omissions. Some of these, but not all, are made
H 26.8
f. "rv aKA7rlpOv Kat d&vTrTvtrov (also K 97. I2 f.). A: 'a
distention that offers resistance to the hand.' The Aragood by being written lengthwise in the margin; so that,
bic appears to omit aoKArlpv Kat (cf., however, H 27.4,had we not the Greek before us, they might be taken for
where two of the three Arabic MSS have 'induration'
glosses. Each probably represents at least one line of the
original Arabic text of the translation.
for aKr pOs).
As far, however, as transcriptions of single words are
H 28.5. There is no equivalent, in the Arabic, for the
concerned, it would be difficult to say that errors are more
Greek Kat otov &vappot8ovp.CLvov [Trpdo rtVo0 7T6v
frequent in one of the transcripts than in the others; but a
hYvrS].
curious feature of the two treatises (Princeton) transcribed
H 28.21 f. ~Lr8' atV7o 80 7d T% PLeXOV wfa 7Erp7l7iEvov EK T2
in A.D. 1726 is that two words, one in the De sectis, and the
;ovae?s og. A: 'even though the body that contains it has
not altered its nature.' There is no trace, in the Arabic,
other in the Ars parva, are correctly given only by this late

euphony.

of an equivalent for -E-rp~/pvov. Moreover the point at

issue appears to be the widening of pores, rather than


the body being pierced or perforated. For the idea of

scribe; a fact which suggests that he may have copied from


very old manuscripts. Since one of the words concerned is

a transliteration of a Greek word KaSpeda [vKr], and the

the body being altered, cf. H 29.9 f., where the point is

other, a word meaning 'the colour indigo', is formed

put thus: 9rS&V at~doCv Tv ow~pcdrbov trapd TO Kada

from the Persian lilanj (the Indigo plant), it seems more


probable that the scribe's success with both is to be attributed to the excellence of the MSS he copied, than to his
own learning.

vownv aAod'orEpov XOdv~wv.

H 28.25 There is no equivalent, in the Arabic, for the


Greek aKpLwtS9.
H 29.26 f. The order in the Arabic is XEpl r'iXEt flpaxtovt
7roSL KV7I-Lq Fp4rpy, i.e. first hand/arm and its parts, then
foot/leg and its parts.

H 30.13 There is no trace, in the Arabic, of dyyEtwv (in


LM, but omitted by H and K o103.8).

J. S. WILKIE

G. E. R. LLOYD

Cambridge

H 30.22 "r AE7Trd'rEpov (with Lm). K o103.16 (with v)


ATrotEpi. A: 'the most tenuous of what is in them'some support for H.
Note on the Arabic manuscripts

For the Arabic, we have worked from two microfilms


representing, respectively, the content of the Manuscrit

2 See M. le Baron de Slane, Catalogue des Manuscrits Arabes (Bibliothb-

que Nationale, D~partement des Manuscrits) Paris 1883-95; Philip K.

Hitti, Nabih Amin Faris, Butrus 'Abd-al-Malik, Descriptive Catalog of the

Garrett Collection of Arabic Manuscripts in the Princeton University Library

(Princeton Oriental Texts v) Princeton 1938.

This content downloaded from 158.251.134.79 on Tue, 20 Sep 2016 16:10:01 UTC
All use subject to http://about.jstor.org/terms