Definition and types of linguistic examples
In linguistics, an example is a linguistic representation of a string of elements instantiating an object language. Its primary purpose is to render an assertion in the context more concrete by providing individual instances for the concepts figuring in it. As such, it must be more readily intelligible than the metalinguistic text which it exemplifies. Whether or not something is intelligible partially depends on conditions in its recipient. In many publication situations, it is not presupposed that the reader knows the object language. If he does not, he has no intuitive access to the example. Therefore, the author must expend considerable effort in the elaboration of the example. This includes not only kinds of information provided with it, but also a reader-friendly layout.
Here the focus is on strings composed of words. Since an example is a linguistic representation, the same distinctions apply to it:
- By its methodological status, an example may be a piece of data or may just be an illustration.
- A probatory example is one which can serve as scientific evidence. To fulfill this function, it must be a piece of data.
- An illustrative example is one which serves as an illustration of an aspect of the linguistic system. It does not need to have the methodological status of data, but may be invented or abstracted from actual data.
- By its position in the context of publication, an example may be part of a text edition or figure as a separate item in a metalinguistic context.
- A contextualized example is part of a text edition displaying a uniform linguistic representation.
- A solitary example is inserted as a piece of an object language in a metalinguistic context.
In scientific contexts, these two subdivisions do not cross-classify completely, but yield only three kinds of examples, since contextualized examples are probatory examples. This is so because a text edition is only done for texts which have the status of primary data. Texts lacking this status include lessons of text books; these are not used in scientific contexts.
The example is composed of a set of representations as explained suo loco. Moreover, it is provided by certain metadata. The linguistic representation and the metadata required differ for the three kinds of examples. Needless to say, the representation levels for the composition of examples depend on the intended readership, in particular their knowledge of the object language and of the metalanguage.
A linguistic text edition follows, in principle, the standards of a diplomatic text edition. It usually differs from the latter in providing additional linguistic representations. The text is commonly presented as follows:
- The metadata comprise identification of the language and the dialect, information on the author or speakers by some appropriate sociolinguistic parameters, date and place of origin of the text, persons involved in the edition.
- The text has a title and an ID which identifies it if examples are extracted from it.
- The running text itself is subdivided into sentences, each of which constitutes a contextualized example.
- The sentences are numbered consecutively.
- Every sentence is put in the same multilineal representation, commonly involving a subset of the levels of the table of linguistic representations:
- one or more representations of its significans
- a representation of its grammatical structure, commonly as a morphological gloss
- one or more representations of its meaning, commonly as translations into the language in which the text is edited (and published) and/or the regionally dominant language.
Specimina of such editions are presented on the website of Yucatec Maya.
If the context deals with more than one language, the language and, if the case may be, the dialect instantiated by the example are identified.1
A probatory example comprises, in principle, the same kinds of data as the contextualized example:
- The source of the data is specified:
- The data is taken out of a corpus, including a text edition: the ID of the text and of the sentence are given.
- The source is published: the metadata required include a bibliographical reference.
- The data is taken directly from oral communication: if allowable, the speaker may be identified.
- Then follow the linguistic representations as before.
An illustrative example is usually less complex in all respects:
- No metadata beyond the identification of the language is needed.
- Depending on the purpose and context, a representation of the significans may be followed by other representations.
Textual layout of the example
A solitary example may either be set off from the running text or be integrated into it.
Example set off from the running text
- Such examples are numbered consecutively throughout a publication; their number is their ID.
- Following the metadata, the example is formatted like a sentence of an edited text, except that its running number in its original text goes into its metadata.
is such an example.
|.||Hittite||(KBo III Rs. 57f + KUB XXVI 71 6f, ap. Lehmann 1984:179)|
|With the booty that I had brought home from the campaign,|
|I adorned them.|
If the layout requires a linebreak, it is good reader-friendly practice to put it at a major syntactic boundary.
Technically, the pair of a string of original text and its interlinear gloss is a table of two rows and one column per word. This, however, is true only technically. Visually and systematically, the example is not a table, but a piece of running text. If the original text takes more than one line, so many separate tables of two rows are produced. Publishers sometimes merge such a set of tables into one, such that each so-manieth word of the second line is in one column with the corresponding word of the first line, and so forth. This renders normal reading impossible, as it makes a piece of running text appear like a table.
Example integrated in the running text
A linguistic example inserted in the running text has the following form:
- Indication of the language precedes the example.
- Longer language names may be abbreviated, e.g. Lith. kalbà ‘language’.
- The textual representation of the example follows. Normally, only a small subset of the linguistic representations is used.
- The quotation of an expression of the object language in the metalinguistic context of the text body is marked off typographically. The appropriate brackets for each of the representation levels are used. Alternatively, an orthographic representation may be put in italics. This is, in fact, the default.
- The interlinear morphemic gloss is enclosed in parentheses.
- The translation into the metalanguage is enclosed in quotation marks.
In sum, such an example may look like this:
In German lief (run\pst) ‘ran’, the past is coded by apophony.
Orthography in example translations
Conventions for the orthographic representation of examples are in another section. Here, a comment on the orthographic representation of the example translation is necessary: Since the translation of an example set off from the running text occupies a line (a paragraph, in the technical sense) of its own, there is a wide-spread desire (esp. among copy editors) to treat it orthographically as a sentence and, consequently, to start it with upper case and to end it with a full stop. The correct treatment, instead, is to use these devices just in case the example is a sentence (or more than one). A major portion of examples figuring in linguistic texts is not, but reduces to a phrase or a dependent clause. There is then no place for initial upper case and a final full stop there.
1 Information on the language and the dialect provided in the metadata of the solitary example is limited to naming them. Further information on their genealogical affiliation or the geographical location of the speech community is provded elsewhere in the description.