The origin of Indo-European ego

(The following uses a special font to render linguistic symbols. Please download and install the Code2000 font [zip] if you're having difficulties with the display. A single asterisk (*) represents reconstructed forms; double asterisks (**) mark implausible forms.)

This blog entry is not about a psycho-linguistic exploration of our collective ego, nor even my own deranged ego. I'm talking about the word 'ego' and its unsaid curiosities, how it relates to Indo-European languages, why Nostraticists are wrong, and where it really comes from. Depending on your personality, my explanation will either prove to be terribly more boring or wonderfully more exciting than you've probably read elsewhere. Bon apétit, mes amis.

The English word 'ego' comes directly from Latin egō 'I, myself', an ancient word used to convey the first person singular throughout the Indo-European family of languages (Old English ic, German ich, Greek égō, Sanskrit áham, Hittite uk, etc.), the common origin being Proto-Indo-European *h₁éǵoh₂ or *h₁éǵom. (Often, simply the common pronominal stem *h₁eǵ- is cited.) Unlike the second person where we see a nominative form *tu, enclitic *twe and plural verb ending *-te-, we see a curious break in the pattern with first person nominative *h₁eǵ-, enclitic *me and 1pp *-me-. Many would expect something like **mu in the nominative instead. However we have to look outside of PIE altogether, in language groups suspected to be remotely related to it (Proto-Uralic *minä and Etruscan mi), before we see any evidence for this earlier pronoun in the nominative case.

Proto-Indo-European (PIE) was spoken around 4000 BCE until the people who spoke it significantly spread out in different directions. Various new languages sprang forth from it, now spoken throughout Europe and India six thousand years later. So we can conclude logically that *h₁eǵ- was first used in Indo-European sometime even before that date.

However, beyond this, intelligent discussion dies. Little of substance is mentioned in books or shared in online forums. Few competent people delve seriously into prehistoric linguistics. Few ask important questions like "Where did that word come form?" or "When exactly was that pronoun first used?" Few too resist idle fantasy when seeking answers to these important questions but we need to solve these questions scientifically, based firmly on facts and logic, not by lazy whim.

Many Nostraticists wrongly connect it to a seemingly identical first person pronoun ɣem in a Siberian language called Chukchi (eg. Joseph Greenberg). This is rejected by mainstream linguists because it relies on idle eyeballing of look-alikes and exposes a lack of knowledge in the languages in question. Most Nostraticists fail to get intimate with the protolanguages on which they write and they suffer much-deserved scorn by academics as a result. However, the same academics who reject this solution offer no real solutions of their own, shirking their duty as scholars to be inquisitive. Sufficed to say, there is a more logical solution already available that shows that this pronoun is a much later invention and unique to PIE.

PIE *h₁éǵ- always seems to have ended, regardless of dialect, in a first person singular suffix intended exclusively for verbs. The ending -oh₂ is specifically used for the indicative mood of so-called thematic verbs (ie. stems ending in *-e-) . So, PIE *bher-e- 'carry' was conjugated in the first person as *bhéroh₂. The ending *-om is likewise a first person ending, used in the thematic subjunctive to convey a hypothetical situation. Either way, this pronoun undeniably behaves as though it were a verb, not a pronoun. Yet, how can a verb be a pronoun?

Easily, actually. It turns out that *h₁éǵoh₂ or *h₁éǵom can be understood to have originally meant 'I (am) here', formed from the attested adverb *h₁e-ǵe "here". This pronoun was initially only necessary when introducing oneself as a new topic, as emphasis, since the verb of a sentence itself was always conjugated distinctly for each person. Numerous languages show similarly formed pronouns made like verbs. Look at Inuit (uvaŋa 'I' from uva- 'here' and -ŋa 'I'; ivvit 'you' with -it 'you'), Aleut (tiŋ 'I' from ti- 'here' and -ŋ 'I'), Coptic (ntok 'you', ntof 'he', ntos 'she' with pronominal endings -k, -f and -s respectively), and even Chukchi (ɣem 'I', ɣet 'you' from a demonstrative stem ɣe- and endings -m 'I' and -t 'you'). Chukchi pronouns may not be related to PIE ones but they do show an independent, parallel development that's still useful here.

The only objection left to analysing the PIE first person pronoun as a verb derived from an adverb is the belief that adverbs aren't normally made into verbs like this in PIE. I encountered that objection once when discussing it online and I didn't know what to say about that until I encountered this informative article entitled Hittite hi-verbs from adverbs that eliminates that argument.

The article explains that a number of Hittite verbs conjugated with its first-person ending -hi are derived directly from adverbs such as āppai 'to be finished' (PIE *h₁ópi 'afterward'), parā- 'to come forth' (PIE *pro 'ahead'), šanna- 'to conceal' (PIE *sn̥h₁- 'without'). That Hittite first person ending, by the way, has already been related by PIE experts to the thematic ending, precisely the one we see in this first person singular pronoun! (See Piotr Gasiorowski's Homepage on IE grammar.)

So my job is done. The mystery of 'ego' is thoroughly solved. Yes, break out the champagne. The mystery is solved. The next mystery now is what the original Pre-IE pronoun that *h₁egoh₂ had replaced looked like...