tgpedersen writes:
> 'jestem' is inflected like 'byl\em', right?
Yes, you're on to something. But, again, that's handbook knowledge. In
early Polish the old forms of 'to be' were most often used with the
l-participle to form the periphrastic past tense. Since they were often
reduced to weak forms (rather like English 'm, 's and 're), they were
increasingly felt to be grammatical particles rather than full verbs.
3pl. <sa,> had no weak form, and weak forms of 3sg. <jest> were very
rare, but both cold be omitted in the 3rd person of the preterite. Thus,
the past tense of 'to be' provided a model for the formation of a new
"strong form" of the present:
sg.[m.] 1. byl/-e(s')m', 2. byl/-es', 3. byl/ (jest)
pl.[m.] 1. byli-smy, 2. byli-s'cie, 3. byli (sa,)
(There was also a dual with person-marking particles added to m.
<byl/a->, f.n. <byle->, but since it went out of use quite early, I'll
skip it.)
Note that the ending of the 1sg. past tense (Modern Polish -(e)m) is
also, historically, the weak form of Old Polish jes'm' < *h1esmi. The
<s'> was still there in the 15th century.
Ideally, the analogical present should have looked like this:
sg. 1. jest-e(s')m', 2. jest-es', 3. jest
pl. 1. sa,-smy, 2. sa,-s'cie, 3. sa,
Such forms of the 1/2pl. indeed coexisted with others in the 15th c. and
later, but were eventually eliminated by their doubly analogical
competitors:
pl. 1. jest-esmy (Mod.Pol. -es'my), 2. jest-es'cie
The conservative 3pl. form might in theory have been replaced by <jest>
(this actually has happened in Russian), but in Polish it has managed to
survive despite the pressure of analogy.
Piotr