Some Latin irregularities explained
by: Fl. Claudius Iulianus
Look at these strange and unpredictable things:
(1) nix 'snow (nom.)' -----------------> nivis 'of the snow (gen.)'
(2) vinc-ere 'win' ----------------------> vic-isse 'have won'
frang-ere 'break' ----------------------> frac-tus 'broken'
(3) genus 'kind, class (nom.)' -------------> generis (gen.)
corpus 'body (nom.) ' --------------------> corporis (gen.)
These things - as other irregularities - in Latin can be explained by looking at reconstruted proto-indoeuropean:
(1) If we look at other indoeuropean languages: English , Greek , Russian , Lituanian we can reconstruct the indoeourpean stem for 'snow' as <*snighw->. An * represents a reconstructed but actually non attested word. Also /ghw/ is a labio-velar phoneme (not tree): it represent an aspirated labialized voiced stop. In germanic languages this stop is reduced to the labiovelar glide /w/, Greek reduces it to a labial aspirated stop /ph/ (this also represented one phoneme, not two!). If we take into consideration the flexive endings:
nominative: <*snighw-s> --------> Ancient Latin <*sni:gs> ---> Classical Latin = [ni:ks]
acussative: <*snighwm> --------> Ancient Latin <*sni:wem> ---> Classical Latin = [ni:ks]
(2) This also can be explained as a residue of the ancient indoeuropean flexion. The -n- also apears in other Indoeuropean languages, such as Sanscrit, in the same position where it apear -n- in Latin verbs!
(3) Rotacism the change of voiceless /s/ into the voiced /r/ that is frequent in languages lacking [z] sounds (for example Popular Madrilian Spanish). Originally the Ancient Latin was <*genes> 'kind, class' and <*corpos>. The first by analogy tended to be realized as the second by Etruscan influence on non-tonic latin /o/ become /u/, if we take in consideration such forms:
corporis < (*corpozis <) Ancient Latin <*corposis> (a perfectly regular form!)
generis < (*genezis <) Ancient Latin <*genesis> (a perfectly regular form!)
In the same maner but a little more complex form can be explained forms such as:
vinc-ere 'win' ---------> vic-isse 'have won' (IE <*win(k)-> 'to win')
frang-ere 'break' -------> frac-tus 'broken' (IE <*bhreg-> 'to break', indeed English < IE <*bhreg->).
(5) The irregularities in : also are reflexes of the ancient flexion of verbs <*som, *ess, *est, *somos, *este, *sont> the /e/ - /o/ alternations also can be seen in English series such as from IE .