BRIEFLY: for pairs of languages there seem to be ‘sound substitutions’ that systematically map words of one language to words of similar meaning in another, something explicable by supposing that the words in question languages are descended from a common ancestor, with systematic and divergent sound evolutions along the way. Working with pencil-and-paper in a fashion akin to cryptography, historical linguistics have over many years proposed the details of many such sound corresponce/evolution systems. This project would aim to put these pencil-and-paper ideas into computational implementations. Amongst other things, such tools could be used to assist in the presentation/verification of the data that historical linguists have assembled to support their proposals, something which could aid people trying to learn historical linguistics
AT GREATER LENGTH:
Sometime during the 19th century Grimm observed that certain sound correspondences seem to recur when comparing words of similar meaning from (old) Greek and (old) Germanic eg.
p~f (pyr/fyr ‘fire’), k~h (kardia/herta ‘heart’), d~t (duo/twa), th~d (thura/dura ‘door’)
and it was later proposed to explain explain examples of this kind by supposing the words to have descended via language specific sound changes from a common origin in an ancestor language eg.
Orig->Greek p->p,k->k,d->d, dh->th, Orig->Old Gmc p->f,k->h,d->t,dh->d
[ the reason for the sound changes to differ between the two languages is assumed to be due to a geographic separation of an ancestral speaker group ]
Comparing English and German there are also certain correspondences eg.
p~f (ship/schiff, help/hilfe, ripe/reif), d~t (ride/reiten, side/seite, need/not), t~s (eat/essen,water/wasser)
explained by supposing the words of English and German to have both derived via language specific sound changes from a common from Old Germanic ancestor language eg.
Old Gmc->English p->p,d->d, t->t, Old Gmc->German p->f,d->t,t->s
Today modern English is thought to be a ‘descendant’ language of an Old Germanic ancestor language, which in its turn is one descendant amongst many others (Greek, Latin, Persian, Sanskrit) of a single ancestor Proto-Indo-European language.
Words with a conjectured common ancestor have come to be called ‘cognates‘, and it is a possibility that an English word has ‘cognates‘ in this technical sense in many other languages.
The aim of the project is put (some of) the intricate and extensive scholarly work concerning the sound evolutions in the ‘Indo-European’ languages (eg. http://en.wikipedia.org/wiki/Indo-European_sound_laws) on a computational footing.
One possible use-case would be to implement tools to assist a newcomer to the area who might be reading through a list of claimed ‘cognate‘ words across a range of languages, and could use a computer program make plain which parts are instantiating the canonical assumptions and which are not, eg.
kardia ~ heorta k~h, d~t
In another rather different visualisation direction, one might take a list of cognate words, such as
English father
Old Frisian feder
Old High German fater
Gothic fadar
Sanskrit pitṛ,
ancient Greek πατήρ,
Latin pater
French père,
Spanish padre,
Tocharian pācar
and place them on a map
Other possibilities might involve cognate discovery and discovery of statistical patterns in the system of sound shifts.