Reconstruction:Proto-Celtic/ɸarextus

Etymology
. Matasović, and Pyysalo after him, believe that the term contains the prefix. Zair on the other hand still believes the identity of the prefix to be unclear. Lindeman on the other hand outright denies the existence of any prefixation on this term. However, everyone agrees that the word contains the suffix.

The palatalization in Goidelic is unexpected; a form would be expected (see also  < ). Reconstructing would yield the Goidelic forms directly but lead to Welsh **eirith and the like due to i-affection, necessitating a reconstruction *ɸarixtā so that the i-affection is blocked by a previous a-affection. In fact, Stifter (personal communication) thinks that *ɸarixtā may have yielded the Goidelic forms too, given how its case-forms are mainly attested in late manuscripts whose scribes were unable to distinguish e.g. -ae (genitive singular of ā-stems) and -a (genitive singular of u-stems) since they had merged in later scribes' speech.

There are multiple theories on the underlying root's identity.
 * Zair mentions a theory by Vendryes where the term is related to 🇨🇬.
 * Lindeman derives this term from a k-extension of . He also identifies this extension in 🇨🇬 and 🇨🇬.
 * Matasović makes a connection with.
 * Pyysalo admits that Matasović's derivation is valid, but offers an alternative connection to 🇨🇬, for which he reconstructs (note that Pyysalo does not believe in three laryngeals, he only believes in one laryngeal pronounced like *h₂; in turn he also believes in the existence of *a and an array of a-diphthongs).

Noun

 * 1) assembly, meeting
 * 2) conversation, discourse