ETCSLhelpSignSign name: GAR3
Values: gar3, gara3, qar


  1. How to search
  2. Updates and changes
  3. Characters and symbols with special meanings or functions
  4. Parts of speech
  5. List of determinatives
  6. List of abbreviations of proper nouns
  7. Browsers
See also Critical apparatus and display conventions.

1. How to search

To search the ETCSL, enter a word, lemma or label in the search box to the right of the Search button and click the button once. If the item is found in the corpus, the result of the search will appear in the form of a KWIC (Key-Word-In-Context) concordance.

Each concordance line starts with a reference to the composition where the line is found. By holding the mouse pointer over this composition number, the name of the composition will appear together with the exact reference to where in the composition the line is. Clicking on this reference will take you to the transliteration, where the concordance line can be seen in context.

Next to the number is a link symbolised by the letters Tr. This link will take you to the translation of the relevant passage. Next to the Tr link is yet another link, Gl. Clicking this will open a new window/tab with glosses, in English, of the individual words in the current line and the lemma form of each word. Each lemma form is in itself a link to a search for that form in the electronic version of the Pennsylvania Sumerian Dictionary.

The drop-down menu immediately to the right of the search box contains a list of all searchable items, e.g. (word) form, lemma, label. A phrase of maximum four words can be searched using regular expressions. For more information about searching, see 3 below.

By default, the concordance lines are sorted by the keyword(s) followed by the first word to the right of the keyword. This can be changed by selecting either right l/f (l=lexeme, f=form), left l/f, or comp. no from the rightmost drop-down menu. See Updates v. 0.6 for more details on sorting.

The in category and or in comp. no(s). options make it possible to restrict the search to a particular category of compositions or individual compositions, e.g. category 'Proverbs c.6' (all compositions starting with the number c.6) or 'c.', if you want to restrict the search to an individual composition.

Encode char. as lets you choose between three different character display options. The 'ĝ' (g^) and 'g̃' (g~) options require a Unicode-aware browser, while the 'j' option displays all characters without diacritics (cf. 3. below).

The Display option can be set to either 'Crit' or 'Plain' signifying display with or without ('Plain') critical apparatus, that is, without showing missing, supplied and damaged text.

The Full text in new window check box controls whether the full text view of the transliteration or the translation opens in a new window/tab (default) or not. The Gl(oss) will always appear in a new window.

Finally, the Full para. in gloss check box controls the layout of the Gloss view, i.e. whether a paragraph-aligned table should be displayed or not (default).

2. Updates and changes

ETCSLsearch v. 1.3

  • sign names, e.g. CU SAG@g, are now searchable as an Advanced search option; note the use of the symbols . x & + @g @t @s to form compound and complex sign names, e.g. write DAG.KISIM5xU2+GIR2 for 'kisim'; note also the slightly different kwic display with all hyphens removed
  • two new categories (groups) of compositions have been introduced and can be searched as sub-corpora: LCB (Letter Collection B) and HF14 (House F 14); the latter is not complete
  • a paragraph-aligned display has been introduced in the Gloss window
  • the layout of the start page has been changed slightly to accommodate new options

ETCSLsearch v. 1.2

  • two new categories (groups) of compositions have been introduced and can be searched as sub-corpora: Decad and Tetrad
  • clicking on a sign in the glossing window performs a search for that sign in ePSD

ETCSLsearch v. 1.1

  • to limit a search to one lexeme (lemma+label or lemma+part of speech+label), search strings of the following kind can be entered: 'l=cen a="type of vessel"' or 'l=cen p=N a="type of vessel"'
  • clicking on a lemma in the glossing window performs a search for that lemma in ePSD

ETCSLsearch v. 1.0

  • in addition to form, lemma, pos and label, Emesal bases are now searchable
  • multiword labels must be enclosed by "", e.g. '"type of vessel"'; alternatively, you may use the wildcard character, e.g. '%vessel'
  • searching the translations is possible
  • display the current line with lemmas and labels by clicking on the Gl link

ETCSLsearch v. 0.7

  • all word forms recognised as proper nouns have been lemmatised

ETCSLsearch v. 0.6

  • the concordance lines can be sorted by keyword; by the lexeme or form either to the right or left of the keyword; or by composition number; determinatives are ignored when sorting; 'š' is sorted after 's', and 'ĝ' (g^) after 'g'

ETCSLsearch v. 0.5

  • the maximum number of KWIC lines displayed is 2,500

ETCSLsearch v. 0.4

  • searching can be restricted to individual compositions or categories of compositions, e.g. all proverbs (c.6*)
  • when the New window option is checked, clicking on the link to the composite text or the translation in the KWIC view opens a new browser window

ETCSLsearch v. 0.3

  • links to individual compositions in the catalogue files activated

ETCSLsearch v. 0.2

  • link to compositions and translation in KWIC line

ETCSLsearch v. 0.1

  • search for words and lemmas using truncation and regular expressions
  • search can be restricted to Emesal or Emegir
  • choose between Crit and Plain view of the search result
  • choose between different encodings of the non-ISOLatin1 characters

3. Characters and symbols with special meanings or functions

  • % = truncation, e.g. 'lugal%' finds all instances of 'lugal', 'lugal-ka', 'lugal-ju10', etc., while '%lugal' finds all instances of 'lugal' and e.g. 'nam-lugal'
  • _ (underscore) matches one character, e.g. 'mu-ni-in-r_' will find 'mu-ni-in-ri' and 'mu-ni-in-ra'. '______' will find words consisting of six letters, including hyphen.
  • having any of the characters ^, $, [, ] or * in the search string will invoke matching using extended regular expressions. ^ matches start of search string, $ matches end of search string, [ ] enclose character classes, e.g. [0-9] or [aeiu]. '.*' match any character zero or more times, while '.+' match one or more times.
  • using upper case letters forces a search for words with upper case letters only, e.g. searching for 'DU' will find strings containing 'DU', but not 'du'
  • the opposite restriction is invoked by \B, i.e. only lower case instances of the search string will be displayed
  • | (vertical bar = logical OR) can be used to search for more than one string at a time. '&d;utu | utu' will find 'utu' with and without the determinative &d;
  • the following character conventions have been used:
    J = Ĝ
    j = ĝ
    H = Ḫ
    h = ḫ
    C = Š
    c = š

    use the unaccented characters to the left of the equal signs when searching

4. Parts of speech

All recognisable words in the corpus have been given a part-of-speech (pos) attribute. The value of the attribute is an abbreviation, e.g. 'N' for noun, and follows the recommendations of the Expert Advisory Group on Language Engineering Standards (EAGLES). In addition to the obligatory values listed by EAGLES, some recommended values have also been adopted. These are the values of an optional type attribute in the <w> tag. Further information about obligatory and recommended attributes can be found on the EAGLES pages. Below is a list of the values of the pos and type attributes found in the corpus. A detailed discussion of the markup of the ETCSL can be found in the manual.

Part-of-speech information
Pos valuePart of speechType valuesExample/Comment
AJadjectivekug, gal-an-zu
AVadverba-da-al, ga-nam, me-a
Cconjunctiontukum-bi, u3
Iinterjectione, u-u8-a
Nnounideophone, DN, EN, GN, MN, ON, PN, RN, SN, TN, WNThe upper case values are types of proper nouns. Common nouns are not marked as such. Example of ideophone: dum-dam
NUnumeralcardinal, fraction, ordinal1, 1/2, 2-kam-ma
PDpronoun/determinerdemonstrative, indefinite, interrogative, nominal-relative, personal, reflexivere, na-me, a-ba, za, ni2-bi

5. List of determinatives

  • &ance; -- before words denoting equids
  • &d; -- before divine names
  • &dug; -- before words for vessels
  • &e2; -- before words for buildings and rooms
  • &f; -- before women's names and words denoting women's occupations
  • &gi; -- before words for items made of reed
  • &gud; -- before words for bovids
  • &jic; -- before words for items made of wood
  • &id2; -- before river names
  • &iku; -- after words denoting surface measures
  • &im; -- before words for items made of clay
  • &kac; -- before types of alcohol
  • &ki; -- after place names
  • &ku6; -- after words for fish
  • &kur; -- before words for countries and mountains
  • &kuc; -- before words for items made of leather
  • &lu2; -- before gentilics and words denoting men's occupations
  • &m; -- before (men's) names
  • &mu; -- before words for items made of wood (Emesal)
  • &mucen; -- after names of birds
  • &mul; -- before names of stars and planets
  • &na4; -- before words for stones
  • &ninda; -- before words for bread and other baked items
  • &sa; -- before words for braided items
  • &sar; -- after words for vegetables
  • &cah2; -- before words for denoting varieties of pigs
  • &tug2; -- before words for garments
  • &tum9; -- before words for winds
  • &u2; -- before words for plants
  • &udu; -- before words denoting varieties of sheep and goats
  • &urud; -- before words for items of bronze and copper
  • &uzu; -- before words for parts of the body
  • &zabar; -- before words for items of bronze

From J. Huehnergard, 2000, A Grammar of Akkadian and W. von Soden, 1985, Akkadisches Handwörterbuch. Band I. Supplemented by Graham Cunningham.

6. List of abbreviations of proper nouns

  • DN = divine name
  • EN = ethnic name
  • GN = geographical name
  • MN = month name
  • ON = other name
  • PN = personal name
  • RN = royal name
  • SN = settlement name
  • TN = temple name
  • WN = watercourse name

7. Browsers

  • your browser (and operating system) should support Unicode
  • works best with Mozilla FireFox or Opera on Windows platforms
  • Mozilla FireFox and Safari are acceptable Mac OS X alternatives
  • other browsers/platforms may not display the diacritics properly

Sumerian scribe

© Copyright 2003, 2004, 2005, 2006 The ETCSL project, Faculty of Oriental Studies, University of Oxford
Updated 2006-03-07 by JE

University of Oxford