<< . .

( 33)

. . >>

subcategories are frequently re¬‚ected in the wording of individual entries.15
Many of those entries have a form something like ˜for the same™, ˜another for
the same™, or ˜also for the same, but made of multiple ingredients™. Whatever
the precise form, the terms alius (˜another™, referring to the treatments) and
idem and item (˜the same™ and ˜also™, referring to the complaint) nearly
always signal the continuation of a group, as might be expected just from
the meanings of the individual terms. However, the groups de¬ned by this
wording in the TOC do not cross the lines between groups in the text
(as de¬ned by lemmata), and in several cases correspond exactly to them.
Subdivision, legible in both the TOC and the main text, further increases
the reference value of the work to the reader. The overall ordering does
the same as well. From the point of view of the user, the important issue
will ordinarily have been to be able to ¬nd the treatment for particular
diseases as they actually present themselves. Hence, that becomes the main
parameter of arrangement. (Within groups, there is a strong tendency to
start with the simplest treatments, which might be considered a similar
There is one set of exceptions to the signalling function of ˜the same™/˜also™
(item/idem) noted above, and it does point to one limitation of Scribonius™
organisational scheme. Each is used a few times to signal that a given
treatment is effective for more than one ailment.16 It has no good way to
deal with such treatments. Now, in a modern drug manual this could be
catastrophic. In that case, where there are more categories of information to
deal with (contraindications, drug interactions, side effects), it is important
to be able to gather the whole pharmacopia together to optimise treatment.
Yet since all of Scribonius™ compositions are equivalently effective (or are
treated as such), the ancient physician was little handicapped. At worst, his
list of potential treatments for any particular malady came in slightly short,
and that only if he did not have time to think through his entire repertoire
(rather than just this speci¬c section of the text). And over the course of
extended treatment, he did not need to depend so much on the TOC.

pli n y
The ¬rst book of the Elder Pliny™s Natural history consists of a table
of contents, listing the various contents in order and divided by books.
(When referring to the text of the TOC below, I will cite the entry

15 There are apparently lemmata between sections 18/19, 31/32 and 200/1.
16 Numbers 1, 25, 26, 81“2, 92, 121, 126, 213.
94 a ndrew m. ri g g sby
corresponding, say, to book 5, chapter 12 simply as 5.12. I trust context will
disambiguate chapters of the main text from their TOC entries.) It is pre-
ceded by the preface, so the work has essentially the same structure as Scri-
bonius™. There is also a fair amount of non-summary material interleaved
into the TOC. Pliny offers counts of the ˜facts™ recorded and lists of the
authorities used in each book. None of this information is correlated with
the topics.
I need to preface further discussion of Pliny by saying a few words about
the chapter numeration of modern editions. By working back and forth
between the TOC and the main text, it is generally possible to divide the
latter into segments that must have been intended by Pliny. The list of topics
and their actual execution generally line up, and dif¬cult cases can often be
resolved by establishing the borders of surrounding chapters. There are a
few problems with the modern numeration. In some cases a single chapter
is made to correspond to what look like two topics in Pliny™s index (e.g.,
8.61). And in some books (e.g., 19, 28) there are long swathes of numbered
sections that do not correspond to anything in the TOC. This seems to
indicate a lapse in Pliny™s TOC. Most reconstructible chapters are of fairly
modest length, and there are seemingly obvious changes of subject within
the potentially unbroken sections of book 19. There are also occasionally
problems of ordering in the TOC. So, for book 25 we read:
on the greatest pain (25.7)
discoverers of noble plants (7“39)
moly III (8)

At ¬rst, it might seem that editors should have divided what they call
chapter 7, but the discussion of discoverers comes up before that of pain
in the chapter. Nonetheless, it seems fair to take these chapters as original
units, at least compositionally, unless there are speci¬c reasons not to in
individual cases.
In the process of lining up TOC and text, modern editors have seen that
some entries in the TOC seem to indicate higher order categories than the
rest. That is, they correspond not to a single chapter, but to a whole set
of related chapters, and so ranges of chapter numbers are recorded next
to them. Mayhoff™s Teubner text further emphasises such groupings by
indenting the sub-entries; Beaujeu™s Bud´ prints the higher order entries
in bold-face. The close agreement between editions as to these groupings
seems to indicate something originally Plinian, just as in the basic chapter
division. Take, for instance, this segment of the TOC for book 7:
Guides to the wor(l)d 95
examples of the greatest reverence (7.36)
those outstanding in the arts (37“39)
astrology, medicine (37)
geometry, architecture (38)
painting, sculpture in bronze, in marble, in ivory, carving (39)17
Someone reading the previous sections will note that excellentes here is
substantive: ˜those outstanding in™. The substance (and grammatical case)
of the entries for the next three sections pick up on artibus (˜in the arts™)
and that binds the grouping together. This ends with the new substance
and syntax of the entry for 40. I take it these editorial decisions can also
largely be taken for granted unless speci¬c circumstances warrant. So, for
instance, Naas18 is justi¬ed in taking them as evidence for a greater degree
of structure in Pliny™s working methods than some have wished to concede.
I noted in the introduction that such multi-level structure would be one
of my main interests in this paper, but what I am particularly interested in
is a more audience-oriented approach. Would such a structure have been
useful or even visible to an ancient reader of Pliny? In fact, we will see that
there are numerous problems.
First, there are actually two slightly different situations which modern
editors treat identically. In the passage just cited, the heading artibus excel-
lentes (˜those outstanding in the arts™) is purely conceptual; it can only be
seen as an organisational device. There is no such topic in the main text.
But there are many examples of the following sort:
wonders of the sea (2.101“5)
what powers the moon has over land and sea (102)
what the sun (103)
why the sea is salty (104)
where the sea is deepest (105)
Here ˜wonders of the sea™ could refer to a single chapter, rather than a whole
grouping. The two con¬gurations are roughly equally common.19 And in
a few cases of this sort, the opening section of a group in the main text is
itself nearly just a lemma:
17 summae pietatis exempla (7.36)
artibus excellentes (37“39)
astrologia, medicina (37)
geometria, architectura (38)
pictura, scalptura aeraria, marmoraria, eboraria, caelatura (39)
18 Naas (2002) 189.
19 Ignoring groupings of only two chapters, I count 43 where the purported super-category also corre-
sponds to a chapter and 53 where it was purely organisational. Editorial differences might make for
small changes in these ¬gures, but would not affect the general picture.
96 a ndrew m. ri g g sby
wonders of ¬re and water together (2.107“110)
on maltha (108)
on naphtha (109)
places always on ¬re (110)
The segment of the main text that could correspond to the proposed lemma
(˜wonders of ¬re and water together™) reads: ˜Let us add certain wonders of
¬re (the fourth element in nature), but ¬rst some from water™ (107).
Now, frankly, it would have been hard for a reader to see much of this
without use of modern typographic conventions, such as the indentations I
have used above. Similar devices were not unknown by Pliny™s time (notably
in inscribed legal texts), but there is no evidence of any papyrus text relying
so heavily on them.20 (And note that with something like the Bud´™s bold-
face headers, one can ¬nd the ends of sections only by reference to printed
numbers, which seem clearly not to have been original.) This combined
with the underlying formal inconsistency (pure lemmata in TOC or main
text or neither) may suggest that what structure there was in the Natural
history was not for the direct bene¬t of the reader.21 The same is suggested
more strongly by some other devices that could have been used to structure
the TOC. There are entries in the TOC which are introduced with terms
like ˜their™ (eius/eorum), ˜in which™ (in quibus) or ˜in the same way™ (item)
that could point out groupings. For instance, Pliny™s famous treatment of
elephants is recorded as follows:
On elephants (8.1“11)
on their senses (1)
when they were ¬rst harnessed (2)
on their docility (3)
wonders among their deeds (4)
on the nature of beasts in grasping their own danger (5)
when elephants were ¬rst seen in Italy (6)
their combats (7)
how they are captured (8)
how they are trained (9)
on their birth and the other matters (10)
when they are born. Their enmity with dragons (11).22

20 Moatti (1997) 222“3.
21 The same is also true at a higher level at which Pliny gives some books, but not all, individual titles.
22 de elephantis (1“11)
de sensu eorum (1)
quando primum iuncti (2)
de docilitate eorum (3)
Guides to the wor(l)d 97
This passage illustrates a number of potential strategies. The frequent use
of ˜their™ (eorum) tells the reader to refer back, suggesting a continuing
sequence. Alternatively, one might repeat a key term to emphasise ongoing
sequences, as here at (6), but also more extensively at, say, 13.24“6. Terms
meaning ˜the rest™ like cetera (e.g., 6.5) or reliqua (here) not only suggest
membership in a sub-section, but also the end of that section;23 here, how-
ever, Pliny has placed reliqua (10) somewhat sloppily. More abstractly, one
might note the ellipsis of the subject (in all cases = ˜elephants™) in 2, 8, 9 and
11. Yet all of these strategies are used only intermittently. The segmentation
they do create does not seem to ¬ght against the one reconstructed on a
purely topical basis, but it almost never adds anything either.
The issue of ellipsis may be of further interest here. This would be an
extremely subtle signal if it were meant for bene¬t of the reader. Instead,
it seems more like a symptom of the composition of the TOC based on
the linear text of the Natural history. To repeat, I do not wish to deny
that Pliny™s composition involved structural sophistication; I suggest that
it is not always obvious in the work itself (nor necessarily even in his
ultimate outline).24 If Pliny indeed composed the TOC in this fairly close-
to-the-ground fashion, then he could easily have lost much of his own
structure. This would also explain a phenomenon such as the following.
At 8.17“21 we are given the heading ˜on lions™, followed by ¬ve reasonably
straightforward sub-headings. However, if one reads chapter 8.17, it actually
introduces not only lions, but the other big cats discussed in the following
chapters (panthers and tigers). I noted above that the reader would have
dif¬culty knowing how long a section might go on even with knowledge
of when it started. If Pliny were composing the TOC linearly and close
to the ground, he could well have the same confusion, and that would
explain why the lemma for 8.17 is too limited. In such a large work as
Pliny™s Natural history, a multi-level hierarchy would have been valuable
both for reference and for segmentation. Pliny may have had some of this

mirabilia in factis eorum (4)
de natura ferarum ad pericula sua intellegenda (5)
quando primum in Italia visi elephanti (6)
pugnae eorum (7)
quibus modis capiantur (8)
quibus domentur (9)
de partu eorum et reliqua natura (10)
ubi nascantur. discordia eorum et draconum (11)
23 See already on these terms Doody (2001) 22, n. 20.
24 Naas (2002) 192“3 argues that some numerical information in the TOC is based on an intermediate
text or set of texts, rather than the main ¬nal text. This is possible, though not certain.
98 a ndrew m. ri g g sby
to start with (though perhaps inconsistently), but he passed little of it on
to his reader.
Let me also mention one other organisational oddity here, which will be
taken up more analytically in the following section. Books 29 and 30 both
give animal-derived remedies. Both books, however, include almost exactly
the same set of animals (and animal parts). Those in the former book are
˜medical™ as indicated by the head-matter on the history of medicine; those
in the latter seem ˜magical™ on similar grounds. The order in each book is
roughly parallel and based on the animal source of the remedy. Each rubric
in either TOC lists the number of remedies based on the given animal in
each book individually and then in total.

co lum ell a
Columella™s TOC comes not at the opening of the work, but as a postscript
to his book 11. It is, however, introduced and explained by the immediately
preceding text. Most of the lemmata in the TOC appear in the main
text as well. This seems to have been an original feature of the text, but
it is not signalled in the introduction to the TOC. It is hard to make
generalisations about the entire TOC. The lemmata take on many forms,
including all those attested in any of the other TOCs of this chapter (single
and multiple entries, all syntactic constructions, long and short, use of
pronouns like is and idem), and even show some unique features. Some of
them address the reader in the second person (3.8, 9; 5.1). At the level of the
individual book, however, we may be able to make more of this diversity.
More speci¬cally, the choice of forms for different books seems to suggest a
choice of orientations that the reader might take toward those books. Books
2 through 4 lay out basic procedures for preparing the land (book 2) and
growing vines (books 3 and 4). These seem to be imagined to provide step-
by-step instructions to the reader. The lemmata are primarily in the form
of indirect questions (how? how many? how much?). Key terms such as
vitis (vine) or malleolus (a technical horticultural term usually translated as
˜slip™) are repeated in successive lemmata, so that they may each stand alone
(here Columella is much more consistent than Pliny). They are arranged
roughly in the order they would need to be carried out through the year.
Once the text even points to the sequence:25

25 One might also point to deinde at 2.3, though that is arguably a logical connector, more than a
temporal one here.
Guides to the wor(l)d 99
How meadows are made from ¬elds
How meadows are cultivated once made.

Here the ˜once made™ (facta) in the second line warns the reader that the ˜are
made™ (¬ant) of the ¬rst line needs to be carried out earlier, imagining not
a timeless set of procedures, but a particular execution of them which does
require an order.27 Twice the verb even falls into the second person, suggest-
ing a speci¬c audience.28 Contrast books 6 and 7 on various kinds of farm
animals (as well as 8 on birds). Here the lemmata are almost always in the
form of ˜on™ (de) plus a word or short phrase. They also show some segmen-
tation. The main sections are arranged by type of animal: ˜on horses™, ˜on
mules™. Subtopics to these follow immediately and are frequently marked
by lack of the preposition and/or use of a pronoun (usually ˜their™) to refer
to the section topics. So, ˜their treatments™ (medicinas eorum, viz., horses)
shows both markers (again, when Columella does this at all, he does it
more consistently than Pliny). Now, this distinction between a sequence
of orders and the more encyclopedic approach is not strictly required by
reality. Horses, for instance, could perhaps be treated chronologically by
working though their life cycles. Nonetheless, the formal differences are
clear, and seem to suggest that Columella imagined his reader consulting
different books in different ways.
We might then suggest that Pliny was doing something broadly similar in
his elaborate co-ordination of books 29 and 30. He has certainly expended
more energy on organisation than we have seen previously. Now, he has
not adopted a physician™s work process as his model. If he had, he would
have arrived at something more like Scribonius™ organisation (i.e., based
on symptoms). He may, however, have in mind a reader who is differently
oriented toward these books than toward the others (whether or not they
happen to be the same person). To specify that orientation is more dif¬cult.
One could imagine a ˜scienti¬c™ or ˜theoretical™ reader, who might want to
compare or contrast the domains of magic and medicine, who could use

26 quem ad modum ex arvo prata ¬ant
quem ad modum facta prata colantur
27 For a parallel phenomenon, see Riggsby (2003) 176“7 where I argue that Pliny™s time-deep familiarity
with his own villas vitiates his sense of their spatial order, whereas someone else™s villa is more likely
to be presented linearly. In Columella the question is temporal rather than spatial order.
28 Here again we can see a parallel phenomenon in Pliny; see Riggsby (2003) 175. In de-ordered contexts,
most people tend to fall out because they have no consistent place in the author™s experience. There
are, however, categories of functionally equivalent persons in some cases, and it might be argued
here that the addressee of advice is such a person (or, rather, place-holder for many people).
100 a ndrew m. ri g g sby
the parallel lists as controls. Or it might be useful to a person who has
one method in hand and wants to compare others (this, we noted, was a
potential weakness of Scribonius™ version). At any rate, we should keep in
mind this kind of book-by-book targeting of longer texts.

ge lli u s
Gellius™ chrestomathy, the Attic nights, has the same general structure as
the Compounds and the Natural history: prefatory letter, table of contents
(foreshadowed by the letter), then main text. Though modern texts usually
print the lemmata at the head of the individual sections, the introductory
letter and the practice of the manuscripts do not really support their place-
ment anywhere except in the TOC. In general, the syntax of lemmata in the
four texts is not as standardised as in modern TOCs. As noted above, they
include bare nouns (or noun phrases), prepositional phrases (usually with
de ˜about™), indirect statements and indirect questions. The form of Gellius™
lemmata is distinctive in several respects. First, he is relatively sparing of de
(and the equivalent super); instead, he prefers indirect questions or (most
notably) a variety of constructions dependent on nouns which refer to
types or segments of discourse: ˜words™, ˜account/history™, ˜epigram™, ˜con-
sultation™, ˜conversation™ and so on. And even when he does use the de/super
construction, the ultimate reference is still normally to text29 or to a word30
rather than the thing itself. For example, ˜On the constellation which the
Greeks call Œmaxan and we call septentriones and on the theory and origin
of either word™ (2.21) tells us more about words than constellations. I will
return to the signi¬cance of this fact shortly. Secondly, there is occasionally
more than one lemma per section, such as the tripartite:
That great honours were granted of old to advanced age, and why afterwards the
same was granted to husbands and fathers, and certain things on this point in the
seventh chapter of the lex Julia. (1.3)
In itself, this would prevent the lemma from segmenting the work; that is,
you cannot tell from the lemmata how many sections there are or what their
topics are without referring to the actual text. Some diacritical feature, such
as indentation, line-breaks, or literal rubrication would be required for that
purpose, and we noted above that this would have been unlikely. (Fortu-
nately, the radical changes of subject between sections guarantee that the
standard section divisions are correct.) However, different questions within

29 30
E.g., 2.7, 22, 24; 3.3; 4.7, 18. E.g., 2.21; 4.16, 17.
Guides to the wor(l)d 101
a lemma are always joined by et or -que (˜and™). This never happens between
lemmata. Third, the information in the lemmata is shaped differently from
that in the sections of the main text. A disproportionate number of the
sections open with the name of the source on which Gellius is drawing.
The easiest thing to do would be to reproduce the same pattern in the
lemmata, too, and this is common enough. However, in a number of cases
the name is delayed until the reader has some sense of the substantive topic
(or even omitted altogether).31 So, for instance 15.10 opens:
Plutarch, in the ¬rst of his books ˜On the Soul™ . . . said that many Milesian maidens
have hanged themselves to death.
while the lemma for the same section is:
On the wilful and wondrous death of the Milesian maidens.
Here there may be a functional distinction. In the text, Gellius immediately
con¬rms for the reader the authority and interest of the text in question.
In the lemmata, he more often puts forward the topic, that is, a criterion
by which the average reader is more likely to be searching.
Beyond the structural similarities of all four works under consideration
here, however, Gellius™ work has particularly close ties to Pliny™s. Both, of
course, are large collections of information on a wide variety of topics, but
substantial overlap in the topics of their prefaces has convinced many that
Gellius is speci¬cally responding to Pliny.32 Most important of these themes
for present purposes is a particular economy of time noted by Henderson
and Ker. For instance, Henderson33 shows how Pliny (followed in part by
his like-named nephew) emphasised a need to maximise one™s return on
available time (that is, time not already devoted to negotium) by means
of studium. Ker34 notes an additional constraint that is less well de¬ned,
but thereby the cause of potentially dangerous strategic negotiation. This
is the proper use of night time, that is the ˜borrowing™ of limited amounts
of darkness to supplement the day, without simply inverting the normal,
healthy opposition of night and day. Both Pliny and Gellius respond to
these imperatives with ferocious acts of compression. Students of Pliny
have seen this as a gesture of power in his work; cataloguing the things
of the world claims actual authority over them.35 It seems natural then

31 Switched: e.g., 2.10, 20; 4.9; 5.7 Omitted: e.g., 1.14; 2.6; 3.1, 5, 6, 9, 13; 4.14; 5.17.
32 E.g. Astarita (1993) 20“3 and Holford-Strevens (2003) 165“6 on adversarial/ironic aspects of the
33 34 Ker (2004) 232“6.
Henderson (2002a) 80“4, 88“9.
35 Naas (2002) 421“3, 435“6 with much secondary literature.
102 a ndrew m. ri g g sby
that Pliny™s TOC recapitulates the gesture made by the work as a whole.36
Reduction of the whole world to thirty-seven books asserts control of that
world; reduction of those thirty-seven books to one does the same thing.
(The quanti¬cation of facts and authorities is perhaps then a third-order
compression.) While I have suggested above that the TOC is not entirely
symbolic, its reading in these terms must be at least partially correct.
What I want to suggest here is a parallel reading of Gellius™ TOC. The
terms of Pliny™s lemmata are overwhelmingly things in the world. As I
noted above, Gellius prefers the terms of discourse. Pliny was part of the
state apparatus and wrote notionally for the emperor. His work is part
of the project of empire and accordingly looks outward from centre to
subordinate periphery.37 Gellius™ dedication, if any, is now lost, but remains
of his preface do not appear to address a single, speci¬c reader. While Pliny™s
writing was distinct from his ˜true™ duties, it was still meant to be useful
to the same emperor. Gellius™ is the opposite of duty, being a matter of
recreation (pr.1). Its function is the enhancement of liberal education in
the most general sense (pr.13). His is a private, individual project. Gellius,
then, is attempting to master not the real world, but the world of discourse,
reducing it ¬rst to twenty books of notes, then further to a single book of
lemmata. His location may even be relevant here. Athens could perhaps
stand as the symbolic centre of this world of texts, even as Rome was the
political centre of the material world.38

conclu s i ons
What I have offered above is a provisional sketch of the table of contents
in Latin literature. Though certain themes came up across discussions of
different authors, I hardly developed either a historical narrative or a single
conclusion. But this diversity in its own right suggests certain conclusions

<< . .

( 33)

. . >>

Copyright Design by: Sunlight webdesign