Origin of the verb form "(я) ем"

Question

Can please someone explain the origin of "-м" in "(я) ем".

Why do other verbs have "-у (-ю)" ending in the 1st person singular of present ("я работаю, я учу"), but this verb has "-м"?

Answer 1

First of all, there are some other irregular verbs.

Wikipedia claims that:

Нехарактерную для глаголов I и II спряжений систему окончаний (архаическую) имеют глаголы есть, надоесть, дать, создать (и их приставочные производные: переесть, заесть, сдать, отдать, предать, воссоздать и др.), а также обсценный глагол ссать (сцать) (спряжение на -ы-).

As for why is this so, well, irregular verbs in any language are traces of old grammar constructions, and not all of them can be logically explained.

Even in Old Russian such verbs had been pretty rare:

К нетематическим (V класс) относилось всего пять глаголов: быти, дати, ѣсти ‘кушать’, вѣдѣти ‘знать’ и имати. В 1-ом лице ед. ч. у этих глаголов было окончание –мь (есмь, дамь, ѣмь, вѣмь, имамь), во 2-ом лице ед. ч. – окончание –си (еси, даси, ѣси, вѣси, но имаши); в оcтальных лицах окончания были те же, что и у тематических глаголов.

So, to conclude, this is remnant of -мь ending.

Answer 2

This ending came from Proto-Indo-European athematic indicative singular first person verb ending -mi. The PIE word for "(I) eat" was e̯ed-mi

You can see an example of full PIE verb declension here:
http://pie.allderek.com/doku.php?id=steh%E2%82%82

The same ending gave the English form "(I) am" from the PIE form of the verb e̯es-mi.

Answer 3

Now let's take a look at a broader scope. First, at two ancient IE languages, Sanskrit and Tocharian.

The endings would be the following.

1. for Tocharian (B-dialect, which is the most ancient one):

   yamaskau = I make/do
   yam = I go
   nasam = I am (A-dialect)

2. for Sanskrit:

   लट्लकार (present tense): भवामि [bhavaami] = I am [for someone else's benefit]
   लङ्लकार (past tense): अभवम् [abhavam] = I was [for someone else's benefit]
   लृट्लकार (future tense): भविष्यामि [bhaviShyaami] = I will/shall be [to someone else's benefit]
   विधैलिङ्लकार (optative): भवेयम् (bhaveyam) = I may be [to someone else's benefit]

3. for Hittite:

   esmi = I am
   epmi = I take
   edmi = I eat
   istamasmi = I hear

and the like.

This is a common personal ending for (Proto-)Slavic and some (non) Indo-European 1st. p. sg. present. For example, in Polish the 1st. p. sg. for present tense is a regular paradigm for a certain type of verbs (alongside with the -ę 1st person sg. for verbs of another type), e.g.

być (to be) => ja jestem
czytać (to read) => ja czytam
wiedzieć (to know) => ja wiem

In Bulgarian, which, like most of languages at Balkanian area has no verb infinitives, there is also a regular м-paradigm:

аз съм (I am)
аз обичам (I like/love)
аз искам (I want)
аз отварям (I open [smth])

We might also compare this to Ghaelig grammar, e.g. the Irish

táim (or tá mé) = I am
beannaím = I bless

or with personal prepositions at the same language:

agam = at me/I have
dom = to/for me
fúm = concerning me

and the like.

The Lithuanian, on the other hand, provides, to the best of my knowledge, almost exclusively the vowel 1st. p.sg. present ending paradigm. In general, the ancient languages provide three different paradigms:

1. m-paradigm, present in most of the IE-languages (either synthetically or as a result of later agglutination process, see the Irish example)
2. vowel-paradigm (sometimes combined with alternative glottoral consonants) and
3. n-paradigm (which is regular for modern Finnish and Estonian. e.g.)

olen = I am (same in Estonian)
haluan = I want (Estonian: tahan)
tarvitsen = I need (the Estonian structure is different, so here is another example: ma sõidan = I travel, I ride)
jatkan = I continue (Estonian jätkan)

Cf. with ni for 1st p.sg. in Basque, Mongolian and Finnish (personal form of nouns).