Languages

LitTEL aims to include information for all languages attested prior to 1000 CE. This includes reconstructed languages, as the comparative method provides a powerful tool for extending observable language change many millennia beyond the earliest documented attestation of writing.

 

Languages versus lineages

A major aim of the LitTEL database is to track diachronic change wherever possible. To facilitate this aim, LitTEL tends to interpret language relations in terms of inheritance, and to keep lineages together unless clear taxonomic splits occur.

In other words, while a concept of discrete languages may be meaningful in synchronic terms, LitTEL operates on the premise that it is generally not meaningful in diachronic terms. In situations of normal, continuous transmission, every cohort of language speakers by definition speaks the same language as the previous cohort.

Thus, for instance, Mycenaean Greek, Attic Greek, Koine Greek and Byzantine Greek are grouped under the same 'language' instance, because their mutual relations can meaningfully be understood in terms of relatively linear chronological continuity (at least within the bounds of synchronic dialectal variation). Likewise, Latin is treated as a single language from its most Archaic stage right to the Romance split (where it divides into synchronically distinct languages).

This means, inevitably, that LitTEL languages correspond to multiple Glottocodes. A column in the GitHub dataset keeps track of this.

Where synchronic dialects of individual languages provide conflicting feature values, LitTEL allows for these values to be incorporated into the dataset too.

 

An overview of the LitTEL languages

This index page provides an overview of the languages and dialects currently in the LitTEL dataset. The following information is provided for each language:

• Its ID in LitTEL and its English Name. The full dataset includes alternative names and a list of corresponding Glottocodes.
• Its coordinates - Latitude and Longitude - for the geographical visualisations. For reconstructed languages the best available estimate of its homeland is used.
• Languages are supplied with chronological information, reflecting their Earliest and Latest date of documented attestation. Reconstructed languages are modelled as a point in time (a language as spoken at the moment that two or more lineages diverged)- and consequently have the same 'Earliest' and 'Latest' value.
• Often, the attestation of a language is skewed towards the 'late' end of its range. The Floruit column allows for the entry of a date which represents the chronological 'centre of gravity' of a language's attestation. Where none is given, it defaults to the mean of the 'Earliest' and 'Latest' columns. Individual value observations are assigned the 'Floruit' date by default when no diachronic change is observable.
• The Attestation status of a language (whether it is documented or reconstructed).

 

A map of the LitTEL languages

This map shows all languages with written attestation prior to 1000 CE. It does not include the proto-languages in the dataset.

Id	Name	Family		Latitude	Longitude	Earliest	Floruit	Latest	Attestation