TLG Technical Note 003:
TLG Citations

Authored: Nishad Prakash, TLG
Maintained by tlg@uci.edu
Created: March 2000
Last Revised: 2000-3-13

The following defines the conventions used by the TLG to assign unique citation values to each line of text in the TLG corpus.

1. Preliminaries

The TLG requires that each line of text in its corpus (delimited by a carriage return) have an associated citation. These citations are used to allow users to readily identify the location in the text of search results. Normally, such citations are based on the subdivisions of the text that already appear in the text edition used by the TLG (the source edition). However, the marking up of title and prefatory material is frequently absent in the source. Therefore the TLG follows certain consistent principles in assigning citations to such segments of text.

2. Citation System

TLG citations can have a maximum of five levels. These levels correspond to units of reference in the printed source edition. These are usually structural units of the text; for example, Book, Chapter, Line, Section, Epistle (in a collection of letters), Oration (in a collection of speeches), and so on. In the absence of clearly marked structure in the source edition, the citation system used is Line, if the edition's lineation is continuous or absent, or Page/line, if the lineation restarts each page.

If a work is a commentary on another work, the citations of the commented work may be included in the commentary's citation system. For instance, scholia on Aristophanes may include the (Aristophanean) Play as the topmost citation level, differentiating the scholia on Av. (Aves = The Birds) from the scholia on Nub. (Nubes = The Clouds). A commentary on a Psalm may include the Biblical verse number of the psalm currently being commented on as its topmost citation level.

The citation system for a work may be found in the database version of the canon on the TLG CD ROMs, as the field cit. The levels of the citation system are differentiated by / or //. For example, the database file contains for Cassius Dio's Roman History (0385.001) the citation system Book//chapter/section/line.

Most TLG-compatible software displays the citation level name for each level of a citation; for example, the first line of Cassius Dio's Roman History, 1.1.2.1, is displayed as Book 1 chapter 1 section 2 line 1 at the bottom of the text window in Pandora and Musaios. In TLG Workplace, the citation system for the work can be seen in the Goto Reference dialog box. In SNS, the citation system can be seen in the Open and Read State windows.

Citation levels can do double work; for instance, a level may indicate either the argumentum or the scholion of the work. In that case, a prefix distinguishes between the two roles; e.g. arg 1 (First argumentum) versus sch 1 (first scholion). The citation system for the work indicates this division of labor with a dash: Argumentum-scholion/line.

Similarly, a citation can indicate two levels of information simultaneously; for example, both volume and section (1,2: Volume 1, Section 2), or both epistle and section (1, 2,A, 2,B, 3: Epistle 1; Epistle 2, Section A; Epistle 2, Section B; Epistle 3). Such citation levels are indicated in the citation system with a plus sign: Volume+section/line, Epistle+section/line.

3. Titles

Title lines are typically excluded from the lineation and sectioning of editions, as they are regarded as extraneous to the text proper. In order to assign such lines a unique citation, the TLG uses the letter t.

The title citation appears a level below the level it applies to. For example, if a work has the levels Book//chapter/section/line, the title of a book is tagged as a title at the chapter level: the title of Book 5 is 5.t.1.1 (Book 5, Chapter 'title'), as distinct from the first chapter of Book 5 (5.1.1.1: Book 5, Chapter 1).

The title of an entire work appears at the topmost citation level of the work. For example, the title of the entire work with the citation system Book//chapter/section/line is considered its own 'Book': t.1.1.1 (Book 'title', as distinct from Book 1, Book 2, etc.)

By way of illustration, the opening of Thucydides' History (0003.001), which has the citation system Book//chapter/section/line, has the following citations:

t.1.
              
THOUKUDIDOU
1.t.
                  
ISTORIÔN A
1.1.
(1.)   Thoukudidês Athênaios xunegrapse ton polemon tôn Pelo-
ponnêsiôn kai Athênaiôn, hôs epolemêsan pros allêlous,
arxamenos euthus kathistamenou kai elpisas megan te esesthai
THOUKUDIDOU is considered the title of the entire work; it is therefore distinct from all the Books of the History, and assigned its own 'title' book: t.1.1.1. ISTORIÔN A is the title of the first book, and so is included in the first book: 1.t.1.1. As the title applies to all chapters of the first book, it cannot itself be included in any of the first book's chapters; it is therefore assigned its own 'title' chapter: 1.t.1.1.

The use of t indicates a title that applies to its entire level. If a title applies to only part of its level, the range it applies to is indicated as a preceding or following numeral or range of numerals. Thus, the title of chapters 3-6 of Book 1 would be tagged as 1.3-6t.1.1: Book 1, Chapter "title of Chapters 3-6". This title would be followed normally by 1.3.1.1 (Book 1, Chapter 3), 1.4.1.1 (Book 1, Chapter 4), and so on.

If a title appears at the line level, it will accordingly be tagged as 1.1.1.t (title of section 1, chapter 1, book 1). If that title consists multiple lines, it is desirable to tag each line with a distinct citation, so that the lines may be differentiated. Rather than calling each such line t, t is prefixed or suffixed by the number of the line. For example, the following is the beginning of John Chrysostom Adversus oppugnatores vitae monasticae (2062.003), which has the citation system Volume/page/line (most Chrysostom being taken from Migne):

47.
(319.) 
PROS TOUS POLEMOUNTAS (1t)
TOIS EPI TO MONAZEIN ENAGOUSIN
 (2t)
  a´. Hote ton en Hierosolumois neôn pollôn etôn  (3)
chamai keimenon apo tês makras epanelthontes aichmalôsias
egeirein êthelon tôn Hebraiôn hoi paides, tote dê barba-
 (5)
The two lines of the title of the work cannot both be tagged as Volume 47 Page 319 Line t, since they would thus not be distinct; they are thus tagged as Line 1t and 2t, instead. Note that, since the title does not apply to the entire page or volume of Migne in question, the tag t could not be used in any case; so it is conventionally associated with the line number of the text.

4. Prefatory Material

The prefatory material to a section is tagged distinctly, when it has been excluded from the lineation or sectioning of the edition. There is no consistent label used in the corpus; it includes p, prol (prologue), pr, pro, proem, prooem, pref (preface), hyp (hypothesis), arg (argumentum), and so on.

Proem-like material is tagged similarly to titles. By way of illustration, the proem to a work with the citation system Book//chapter/section/line is equivalent to a Book, and is distinct from Books 1, 2, 3... of the work; it is therefore tagged as Book "proem", pr.1.1.1. Appian's Civil War (0551.017) has the citation system Book/chapter/section/line; at the beginning of Book 1 there is an extensive proem, spanning several sections, before the beginning of chapter 1. The sections (which are numbered independently of the chapters) thus run:

  1. 1.Pr.1.1 (Book 1, Proem, Section 1);
  2. 1.Pr.2.1 (Book 1, Proem, Section 2);
  3. 1.Pr.3.1;
  4. 1.Pr.4.1;
  5. 1.Pr.5.1;
  6. 1.Pr.6.1;
  7. 1.1.7.1 (Book 1, Chapter 1, Section 7).

On occasion, such material is distinguished from the ancient table of contents of a section or work, tagged as pin or pinax. For example, Hero of Alexandria's Definitiones (0559.008: Chapter/section/line) contains a table of contents (pinax.1.1 - pinax.1.160) and a proem (proem.1.1 - proem.1.9) before its first chapter.

There are two other special letters used in citations. l is used (infrequently) to characterise a line as a lemma. It was used in the CD ROM #D Epicurus, for example, as follows (examples from the Letter Fragments, 0537.004):

(36.) 
PROS ATHENAION (t)
&PAP. HERC. 176 5 XXII VOGLIANO (115 Us.):$ (l)
  Athênai-  (1)
ôi de: P o l u a i n o u  p a r a g e -
Fragment 36 has a title line, 36.t, and a lemma line, 36.l, as distinct from the first line of actual text, 36.1.

n is a less well-defined tag; it is used in general to mark lines apart from ensuing lines, if they would otherwise have the same line number, or if the former lines are excluded from the line numbering of the source edition. This is most frequent in biblical commentaries, where the biblical text is to be separated from the ensuing commentary. The following example is from Origen's commentary on John (ed. Blanc) (2042.005)

  Êdê de theon aitômetha sunergêsai dia Christou hêmin
en hagiôi pneumati pros anaptuxin tou en tais lexesin ena-
potethêsaurismenou mustikou nou.
@1
1.16.
(n.)                 
En archêi ên ho logos.
(90.)   Ou monon Hellênes polla phasi sêmaino-
The biblical citation En archêi ên ho logos is excluded from the running sections of the chapter; it is followed by Section 90 of Book 1, Chapter 16, but itself belongs to no section in the printed edition. It is therefore assigned the section n.

The n tag is also used to denote lines containing musical instructions in hymnographical works, and excluded from the hymn's numeration; and in the non-Greek commentary to a Greek work, particularly in fragmentary works. (It has thus taken over the function of l.)

5. Citation Normalization

Citation levels may not begin with any character other than an alphanumeric. In particular, a citation level may not begin with any sort of bracket. Any instances of such brackets in source edition citations are replaced with a postpositioned asterisk; for instance, the section level (2) is represented in TLG Beta code texts as 2*.