This article was downloaded by: [Aston University]

On: 20 January 2014, At: 12:12

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer
Street, London W1T 3JH, UK

Symbolae Osloenses:
Norwegian Journal of
Greek and Latin Studies
Publication details, including
instructions for authors and subscription
information:
http://www.tandfonline.com/loi/sosl20

Dividing Homer: When

and How were the
Iliad and the Odyssey
Divided into Songs?
( continued )
Luigi Enrico Rossi
Published online: 06 Nov 2010.

To cite this article: Luigi Enrico Rossi (2001) Dividing Homer: When and
How were the Iliad and the Odyssey Divided into Songs? ( continued ),
Symbolae Osloenses: Norwegian Journal of Greek and Latin Studies, 76:1,
103-112, DOI: 10.1080/003976701753387969

To link to this article: http://dx.doi.org/10.1080/003976701753387969

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of

all the information (the “Content”) contained in the publications
on our platform. However, Taylor & Francis, our agents, and our
 licensors make no representations or warranties whatsoever as
to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication
are the opinions and views of the authors, and are not the views
of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified
with primary sources of information. Taylor and Francis shall not
be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with,
in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private

Downloaded by [Aston University] at 12:12 20 January 2014

study purposes. Any substantial or systematic reproduction,

redistribution, reselling, loan, sub-licensing, systematic supply,
or distribution in any form to anyone is expressly forbidden.
Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Symbolae Osloenses 76, 2001

SO DEBATE
Dividing Homer: When and How were the ILIAD and the
ODYSSEY Divided into Songs? (continued )

Luigi Enrico Rossi

On the Written Redaction of Archaic Greek Epic Poetry

I am honoured to accept the offer to intervene in this very interesting SO

Downloaded by [Aston University] at 12:12 20 January 2014

Debate (cf. SO 74, 1999, 4–91, and 75, 2000 , 5–23). “In Homeric scholar-
ship, perhaps more than elsewhere, it is a striking fact that what seems to
one scholar strictly logical, almost self-evident, remains unconvincing to oth-
ers”: so Minna Skafte Jensen (MSJ, SO 74, 1999, 23), with a reference to Rose
1997, 170. None of us can be an exception, both when we speak or write to
an audience and when we listen to a paper or read it. I might misunderstand
essential arguments, I might misuse formulations and terms. It is after all a
danger one has to face anyway.1

On the very nature of archaic epic poetry

The title of the debate reveals an intention to cope with a major problem of
Homeric studies, viz. how did one come to the sole or the multiple redac-
tions of an oral text and possibly when. In fact, even in the variety of schol-
arly positions one can hardly deny that both poems as we have them had an
oral pre-history and one or more written redactions as well. One of the major
problems lies precisely in the division into songs, 2 and I should say it was
high time to start with it. The transformation of the free flow of the many
epic narrations, usual in the oral period, into two great poems or books—the
two major narrative units that were chosen among the epic archipelago—
was a task that must have been undertaken at a time when book-status was
sought for even for a literary product that at its origin was not a book at all:
it was oral, continuous and therefore reluctant to any bookish prison, as I
firmly believe and as will become clear from my argument in the following
lines.

1I apologize for frequently quoting some of my own works on the archaic Greek epos (see the
list of Additional references). I owe more than one important suggestion to Lucio Ceccarelli,
Alessandro D’Avenia, Roberto Nicolai, Livio Sbardella.
2 I shall be using ‘song’ (as the title of the debate opportunely suggests) for each of the 48 units
often called ‘books’, whereas with ‘book’ I mean the book form of the poems as we have them,
viz. shaped as two great poems.

103
 SO D EB ATE

I said at the very start that there is hardly a field in which scholars dissent
more than in Homericis. But when one comes to the highly necessary mo-
ment of offering a possible scenario for the way in which things must have
developed in the course of time, then scholars agree with each other much
more than the usual labels would lead us to expect: unitarism, analysis and
oral theory seem to commingle, so that nowadays I personally find it dif-
ficult to put one of the three labels on a so-called solution of the Homeric
problem. Song division is a structuring device that originally was totally un-
foreseen. A distinction of narrative sections is obvious and is in fact even at-
tested through the titles we find in Herodotus, Thucydides, Plato, Aristotle:
it simply responded to the need of the single performance, but let us not
Downloaded by [Aston University] at 12:12 20 January 2014

forget that the song division we have does not always correspond to clear-cut
episodes. Moreover: an isolated song might be a clear-cut episode, but how
and when did song division become a device for structuring the major unit
that we call a poem?
With this outlook in mind I have been reading the most useful recent
history of the question offered by MSJ (5 ff., 23 ff. on oral theory). Not-
withstanding what I have just said on the difficulty of disentangling labels,
I adopt her rough division (7 f.) of recent scholarship into oralism, unitarian-
ism and classical analysis as useful, and I roughly agree with her assignment
of single scholars to one view or the other of the Homeric problem. But
again, when I come to putting a label on my own opinion of the develop-
ment of Greek archaic epic poetry, I think I am of course an oralist if I
consider the original epic production before literacy, viz. before the 8th cen-
tury BC. But if I consider the poems as we have them I can’t help being an
analyst, so many are the inconsistencies that I see in the text, following the
precious treasure of observations made by positivistic research in about two
centuries (I call them ‘analytical scandals’). And finally I become a unitarian
when I consider the two poems transformed into two books on whose rich
and incoherent material some kind of approximate unity has been superim-
posed. In other words: each scholar should change label according to the dif-
ferent object of his investigation: early epic poetry, the poems at an interme-
diate stage, the poems in the bookish form that we have before us.
Research done under these three labels has in fact been useful and success-
ful throughout, so we cannot dispense with the results of any one of them.
However, one thing has been generally ignored, at least in explicit formula-
tions: that we have to distinguish among the different objects—as I said—
viz. original epic poetry, which was of course oral, rich in themes, not meant
to be unitary at all, in its own time actualized in additional or asyndetic nar-
rative; then the intermediate stages, that are hard to determine; and finally
the poems as we have them, viz. the product of one or better of more redac-

104
 Div id in g Hom er

tors who using literacy have tried and partly succeeded in creating some kind
of unity which in itself is far from satisfactory if the basic conception of unity
is our modern one. If the actual object of research is defined and declared,
many of the misunderstandings that are usual in this field can be avoided.
I have recently argued (Rossi forthcoming) for three different types of
unity, taking our modern one as starting point: lack of unity in original epic
poetry,3 gradual increase in unity when literacy became available (from the
8th or 7th century on), and finally an increasingly modern one from the time
the book as an artifact became widespread (from the 5th or 4th century on),
so that the book produced the literary liber poeticus, which became usual in
Alexandrian and Roman literature. In this field, as in others, we have what
Downloaded by [Aston University] at 12:12 20 January 2014

Droysen called ‘modernity in the ancient world’, dealing with the period that
we have since become accustomed to calling Hellenistic.
I shall not single out the many suggestions that come from the lively dis-
cussion of many scholars on MSJ’s views of song division: I shall only append
some reflections of my own, from which it will be clear what I personally
consider as acceptable in statements and counter-statements. In her paper,
MSJ offers a very useful and thorough analysis of beginnings and endings of
songs: it is clear that they show an overall coherence. But let me say that I
cannot understand what she means when she says that song division must
be ‘original’ (11) and ‘genuine’ (14). What do these two terms mean? What
I think they should mean is something one could better express in different
terms. That song division came about not randomly but in a very intentional
and pondered way, is obvious: only, when is such development meant to have
come about? At a great distance—in my opinion—from original and genu-
ine archaic epic practice. It must have happened when original, genuine epic
practice was over: oral practice of epos, the only one that we can call origi-
nal and genuine, had of course no more than a rough isolation of narrative
units or episodes fit for single performances. 4 In the long way between epic
performances and the redaction of the poems as two books, there is a stage
that has often been considered as original, viz. the practice of performance at
festivals that consisted of successive singers who respected a continuous nar-
ration, what was called eƒ j u„podoxh‚w: the first core of the poems as we have
them or even the poems themselves as we have them. But that kind of struc-
turing the performances was preceded by another way of staging the song,
the way we see practised in the poems: not yet competition at festivals, but

3Including of course Hesiod’s Works and Days as well: see Rossi 1997.
4I find it is otiose and therefore useless to measure even approximately the length of perfor-
mances (MSJ 25 f., 30 f., all.).

105
 SO D EB ATE

less ritual entertainment such as we see with Achilles in the Iliad and with
Demodochus, Phemius and Odysseus in the Odyssey.
Free narrative oral performances must have ruled out any kind of overall
systematic regularity, and MSJ has apparently just that way in mind (ex. gr.
25, 27, 74 ff.). But her very useful pointing out the skilful devices of division
(16 f.), such as morning, night, summary, etc., proves nothing else than the
skill of the redactor or redactors. Moreover, I would point out here an ap-
parent paradox: the more regular song divisions appear, as in fact they do
appear, the less original and genuine these divisions must have been. Add
what I think to be an important clue. If we scrutinize the very useful cat-
egories pointed out by MSJ, we see that the structuring into ‘morning’ at
Downloaded by [Aston University] at 12:12 20 January 2014

song beginning (3x Il., 9x Od.) and ‘sunset’ at song end (3x Il., 7x Od.), to
which one has to add ‘somebody is awake’ at song beginning (3x Il., 2x Od.)
and ‘going to bed’ at song end (4x Il., 7x Od.), is frequent enough to appear
meant for a continuous narration, for the structuring of a great poem, and
not so much for single performances. Even more as such I view the frequent
occurrence of ‘summary’ both at song beginning (17x Il., 4 Od.) and at song
end (11x Il., 11x Od.) and finally ‘rounding off ’ at song end (14x Il., 21x Od.).
Of course there are exceptions and inconsistencies: MSJ is fully aware of
them and tries to justify them in one way or the other or simply accepts
them as such: but, on the contrary, I think that one should go the other way
round, justifying the nearly overall dominating regularity which is so con-
trary to an oral culture. The closer structure and unity are to our standards,
the further away from epic fluidity they are.
It is now clear, I hope, that I cannot help thinking of a long oral phase,
which must remain the very core of every oral creed. He who believes that
the song division may reflect anything ‘original’ is simply paying lip service
to oralism.
At this point I declare my sympathy for one of the sub-labels of Homeric
scholarship, the so-called neoanalysis, especially as represented by a scholar
such as Wolfgang Kullmann, who in his important book Die Quellen der Ilias
(1960) recognised the antecedents of the great poems in the Epic Cycle. Of
course not in the form in which we have the fragments of the Cycle itself,5
but as the former mass of songs orally composed and transmitted. I need
to free myself from the opinion that the Iliad as it is was composed by a
single poet, but this does not diminish the importance of what K. stresses,
viz. that the Ilias had a long prehistory: whether one is unitarian or analyti-

5Already Aristarchus held the view that the Cycle was later than the two poems. Wackernagel
1916, 181 ff. saw in the language clear signs of a later redaction.

106
 Div id in g Hom er

cal is not essential, 6 provided one tries to gain an idea of the nature of what
the redactor or the redactors did. Why on earth is a word such as ‘redactor’
(Wilamowitz’ Bearbeiter) so carefully avoided nowadays? When I hear the
word ‘poet’ I feel rather uneasy, so heavy I feel in it the burden of romantic
aesthetics. Do we need to think that the author of a patchwork (needless to
say, of the highest quality) was working as a romantic poet did? Perhaps be-
cause many think that to define his work as patchwork is equal to disparag-
ing it, whereas one should bear in mind that he (or better they) assembled
not only linguistic blocks (formulas) but narrative (and cultural) blocks as
well, what from Arend (1933) on we call ‘typical scenes’. Linguistic and cul-
tural mixture (war, family, burial customs, etc.) pleads for juxtaposition of
Downloaded by [Aston University] at 12:12 20 January 2014

cultural strata in the same way as refinement in verse rhythm (Hermann’s

bridge and avoidance of mid-verse caesura) pleads for a long development
in time. He who put the whole thing together found much that was already
there as composed poetry.

The transcription
One view I can hardly accept is that one single scribe (or two or three) should
have written down the poems—by dictation that could have been either ex-
ternal or internal—on the occasion of one or more definite performances at
a festival. Historism has taught us to envisage graduality and I am sure that
we have here a typical case. Between the rise of literacy (8th century) and its
full exploitation (in the so-called lyric age, from the 7th century on) there
was room enough for a period of gradual writing down that must exempt
us from too tight precisions. And between the first (gradual) redactions and,
for instance, the Pisistratean one in the 6th century7 we have room enough
for different geographical sites, for more scribes and even poets, for gradual
stages of a growing literary structure. 8
I believe that recently I have been lucky enough to detect a different redac-
tional stage in a parallel epic tradition, Hesiod’s Works and Days (Rossi 1997).
I have identified some repetitions, recognizable through identity of themes
and through traces of more than one incipit : viz. two winters (493–523,
524–563), two navigations (618–645, 646–694), besides other minor ones. I

6 Kullmann himself (1960, 360 ff.) lists three unitarian and three analytical positions that can
share and utilize his achievements.
7 Cassio (forthcoming) argues with solid linguistic evidence for a redaction in Pisistratus’
time.
8 The question of the 24 letters of the alphabet has been abundantly dealt with in the present
debate with pros and contras as far as its antiquity is concerned. Anyway, let us not forget that
Livius Andronicus in the second half of the 3rd century wrote his Odyssey without book divi-
sion, which was introduced later by the grammarians.
107
 SO D EB ATE

have called them ‘alternative sections’, in the sense that they must have been
thematic variants intended for different performances, and certainly com-
posed by Hesiod himself; the same applies to the two Prometheus sections
in the Theogony and in the Works and Days. Whoever assembled and wrote
down the materials that formed the Works and Days and the Theogony 9 as we
have them—Hesiod himself or somebody else—juxtaposed more than one
version of the same subject (meant for different performances), so as to pro-
duce a patchwork, precious for us, which witnesses an intermediate stage in
the long way from original epos to an approximately unitarian book.
I have then compared Works and Days with the Iliad and the Odyssey. In
the great Homeric poems we find no repetitions of narrative units, viz. no al-
Downloaded by [Aston University] at 12:12 20 January 2014

ternative sections, as clearly stated first by D.B. Monro (1901, 325), so that we
usually speak of ‘Monro’s law’. I am fairly confident this means that in the
two poems we have a redactional stage that is more developed than the one
we have in Hesiod. This does not imply that Hesiod is older than Homer; on
the contrary, it confirms his later chronology (7th century): the time that had
elapsed between composition and redaction was much shorter, since literacy
was there at hand; the Homeric poems, on the contrary, had a much longer
time in between so that they could reach a more developed stage of redac-
tion, a redaction without alternative sections. Literacy was introduced when
epic oral practice was still in its heyday, in the 8th century, and the process
of using it for Homeric poetry must have been slower, whereas Hesiod found
literacy already available at the outset and so he offers us the welcome op-
portunity to compare two different epic redactions: his own, quicker and less
accurate, and the Homeric one, which through comparison reveals a longer
elaboration with more accurate results (Monro’s law).10
The succession of different phases of redaction was in fact the consequence
of original epic practice and of its very nature. The Homeric redaction must
have begun as a slow process out of the live epic practice of performance.
The oral culture of archaic Greek epos has been recognised, described, ac-
cepted by practically all scholars in recent decades and it must be our starting
point. Of course it has always been mine as well: my Homeric creed I offered
a long time ago.11 Later I have tried to summarise my view of the problem,
suggesting three different kinds of challenge and maintaining that all three
have been defeated (Rossi 1994): the challenge of authorship, lost because
9 Of course the case of the Theogony is much more complicated: I hope to cope with it in the
near future.
10 The idea of the ‘curative action’ of a tradition (more relevant if the tradition is long) first
came to me from G.S. Kirk commenting on the Homeric Hymn to Apollo (v. 464): Rossi 1981,
221.
11 Rossi 1978; see also Rossi 1995, 23–67.

108
 Div id in g Hom er

epic poetry is the product not of a single personality but of a whole culture,
as Gianbattista Vico very clearly saw in the 18th century; the challenge of
authenticity, lost because it is a consequence of the first one and because,
even if we detect different chronological strata, we have to accept the final
product as we would do with the water of a river next to its mouth and as it
was accepted by all those who introduced their own new cultural and poetic
materials respecting what was already there; and finally the challenge of unity
in our sense, which I have already briefly dealt with above.

Some remarks on early epic practice

Here I shall append a few new remarks of my own concerning the habit of
Downloaded by [Aston University] at 12:12 20 January 2014

focusing on the single episode, so that an overall coverage of the longer story
(Iliad, Odyssey, entire cycles) was something that came only later, at the mo-
ment of successive redactions, when the creative stage was mainly behind.
In my opinion, by far the most important word for Homeric poetics is at
the very start of the Odyssey (Od. 1.10): tv‚n a„mo€yen ge, yea€, yu€gater Dio€w,
eiƒ pe kai h„mi‚n (‘Muse, begin to tell us the tale from any point of the story’).
The epic narration is precisely a ‘cycle’, with practically no beginning and no
end: one has simply to establish from where to begin the performance. When
Odysseus invites Demodochus to sing of the Trojan horse, he begins the
narration (Od. 8. 500) e…nyen e„lvn v„w oi„ men eƒ usse€lmvn eƒ pi nhv‚n ba€ntew
aƒpe€pleion..., viz. ‘beginning from the point when on the ships they sailed
away’. The main thing is to establish where from one starts, since one can
start from anywhere, the whole being a cycle, the Epic Cycle. And further:
how does Odysseus introduce his apologoi? In Od. 9.14 he says ti€ prv‚to€n
toi e…peita, ti€ d’ u„sta€tion katale€jv; It must have been a usual introductory
formula.12
Traditional epos lives through performance, which is every time bound to
a specific occasion, viz. the performance itself of that day. Nothing is more
alien to epos than the fixity of an unchangeable text. I have often wondered
what kai h„mi‚n (‘to us as well’) might mean in Od. 1.10. I do not see any
sense in ‘us’ meaning ‘mankind’ as a polar opposite of gods: the Muse sings
to men. In my opinion it means ‘tell the tale to us as well, here and tonight,
as you have told it to many others and elsewhere in the past’: the single per-
formance is bound to the hic et nunc. And so is the theme of song, since
it can be a quite new one, that is even more appreciated by the listeners:
Od. 1.351 f. thn gar aƒoidhn ma‚llon eƒ piklei€ousƒ a…nyrvpoi h† tiw aƒkouo€ntessi

12 Cf. Il. 5.703, 11.299, 16.692 for the topical list of the killed warriors: e…nya ti€na prv‚ton, ti€na
dƒ u†staton eƒ jena€rijan -jen, -jaw .

109
 SO D EB ATE

ne vta€th aƒmfipe€lhtai, so that the narrative can be steadily added to through

‘news from the world’. Epic poetry in its original phase is an ‘open book’.
Moreover, I should say that even the proems of both poems are not the
proem of each poem, but an occasional one, aimed at a single performance
but then chosen as the proem of the poem. Otherwise the proem of the Iliad
would be clearer and more informative on the subject of the poem as it is: 13
if we had only those few verses we would think that a Greek warrior named
Achilles went mad, armed his troops and went fighting against the Greeks,
something similar to the individual deed of Ajax. The same applies to the
proem of the Odyssey that in announcing Odysseus’s nostos singles out one
episode only, the killing of the Sun’s cattle. How very different the proems of
Downloaded by [Aston University] at 12:12 20 January 2014

the Argonautica and of the Aeneid: they were meant to introduce that poem,
which unlike archaic epos was thought of and built as a unitary book. They
served their aim in a very thorough, simple and clear way. Proems were origi-
nally the alternative section par excellence, which is clear through the alterna-
tive proems that are attested for the Iliad 14 and after all the Homeric Hymns
are simply a collection of proems ready for use in single performances.
All in all: the overall narration made out of the many episodes, the poems
as two books, cannot have been actualized all of a sudden in the still creative
epoch and occasioned by an actual performance at a festival or by the com-
mission of a tyrant. It must have been a gradual process and the division in
songs came about when this process was at its end, when literacy and later
the book as an artifact won the long war against true epic oral practice. That’s
why I find it difficult to think of one or two or so and so many redactors or
poets (MSJ 22, 33 ff.): I think that the one responsible for the actual version
of the poems was the last redactor, who and whose milieu were influential
enough to impose that text as the text of the poems. Those who still stick to
the idea of a strong poetic personality will have to discard much of the evi-
dence we have for more than one redactor coping with great blocks of mate-
rial which they might have added to later with the help of literacy.
Of course adaptation to literacy leads to the belief that there must have
been written additions, which is what MSJ herself (24, 27) believes: she
speaks of “oral or written or something in between” and of “gradual ‘textu-
alization’ over the centuries” up to Hellenistic times. I believe this to be an
interesting starting point for new investigations. 15
13 I owe this observation to Renehan 1987, 115: he is very witty indeed when he suggests what
the proem could mean to somebody who would read only those few verses. The same for the
proem of the Odyssey (Rossi 1997, 17–19).
14 Rossi 1997, 16 ff. (see Montanari 1979, 43–56). Even the proem of the Works and Days is a
proem.
15 MSJ quotes as adherents to this view, apart from G. Murray, also Nagy 1996a, 1996b, de Vet
1996 and myself: I had first suggested it in 1971, but I did not know that while I was writing,
110
 Div id in g Hom er

If after this we try, as we have to, to determine when this last redaction,
structured into single songs, was actualized, we might wonder whether it was
before Pisistratus, or through his commission, or after him, and we might
choose one hypothesis or another. What I think we cannot do is to ignore
that the Iliad and the Odyssey are quite a long way from original archaic
epos and that much of the genuine nature of their origin is not mirrored in
their written redaction: that’s why I suggest calling them simply a witness of
oral poetry, not a document (Rossi 1978). From a broader literary-historical
point of view I should say that the two great poems as we have them, viz. as
two books, are the greater betrayal of a literary genre, viz. of original archaic
epic poetry. The formalists used to call this ‘metamorphosis of the forms and
Downloaded by [Aston University] at 12:12 20 January 2014

of the functions’: new structuring versus archaic free narrative flow, new hy-
potaxis versus archaic parataxis. To summarize through another paradox, the
two poems were viewed in good faith as two unitarian books by a great liter-
ary critic as late as Aristotle and were revered as a literary model, again in
good faith, by late and fatally unepic poets such as Apollonius and Virgil.

Additional references 16
Cassio, A.C. forthcoming: Archaioi Homerikoi: le più antiche edizioni dell’epica e la filologia
dei rapsodi.
Dihle, A. 1970: Homerprobleme, Opladen.
Kullmann, W. 1960: Die Quellen der Ilias (Troischer Sagenkreis) (Hermes Einzelschriften, 14),
Wiesbaden.
Monro, D.B. (ed.) 1901: Homer, Odyssey. Books XIII–XXIV, London.
Montanari, F. 1979: Studi di filologia omerica, Pisa.
Renehan, R. 1987: “The Heldentod in Homer: One Heroic Ideal”, Classical Philology 82,
99–116.
Rossi, Luigi Enrico 1968: “La fine alessandrina dell’ Odissea e lo zh‚low „Omhriko€w di Apollonio
Rodio”, Rivista di filologia 96, 151–163.
— 1971: “Wesen und Werden der homerischen Formeltechnik”, Göttingische Gelehrte Anzeigen
223, 161–174.
— 1981: “Gli oracoli come documento di improvvisazione”, in: I poemi epici rapsodici non om-
erici e la tradizione orale. Atti del Convegno di Venezia, 28–30 sett. 1977, Padova, 203–230;
— 1994: “L’epica greca fra oralità e scrittura”, in: Reges et proelia. Orizzonti e atteggiamenti
dell’epica antica, Como, 29–43.
— 1995: Letteratura greca, Firenze.
— 1997: “Esiodo, Le opere e i giorni: un nuovo tentativo di analisi”, in: Posthomerica I: Tradi-
zioni omeriche dall’Antichità al Rinascimento, ed. F. Montanari & S. Pittaluga, Genova
(Facoltà di Lettere), 7–22.

Dihle 1970 had just appeared: a valuable and welcome analysis advocating the mixture of oral
and written layers.
16For the main Bibliography, see SO 74, 1999, 84–91.

111
 SO D EB ATE

— forthcoming: “L’unità dell’opera letteraria: gli antichi e noi”, forthcoming from a meeting
in Pisa (7.6.1999).
Wackernagel, J. 1916: Sprachliche Untersuchungen zu Homer, Göttingen.

Università di Roma “La Sapienza”

Dipartimento di Filologia greca e latina
Downloaded by [Aston University] at 12:12 20 January 2014

112