Origins of a Pronoun
In Old English, the nominative singular first person pronoun was ic. The Oxford English Dictionary says that after the Norman conquest, the north of England retained this original pronoun in forms such as ic, icc, ig, hic, ik, yk, ike, and hyc. Southern England, on the other hand, palatised it to create such forms as ich, hich, ych, yche, iche, ih, and ihc. By the 14th century the north started to drop the velar plosive before consonants, giving the form i, and by the 15th century this was used in front of vowels too, along with the variant forms hi, j, e, y, Y, and I. The south persisted with ich.
There's a joke about this dialectal division between north and south in the Second Shepherds' Play, the greatest of the Wakefield Cycle of mystery plays from the 15th century, wherein a northern thief called Mak tries to steal some sheep from shepherds Coll, Gyb, and Daw. He attempts to affect a southern accent but fails to keep it consistent:
What! ich be a yoman, I tell you, of the kyng,
The self and the same, sond from a greatt lordyng,
Ffy on you! Goyth hence
Out of my presence!
I must haue reuerence.
Why, who be ich?
Bot, Mak, is that sothe?
Now take outt that sothren tothe,
And sett in a torde!
There is evidence to suggest that this play was still being staged at least as late as 1520, by which time you'd've thought that ich would've been supplanted by I and the joke rendered opaque. In fact, ich held on for a long time. Shakespeare uses some southern dialectal variations in King Lear when he has Edgar say, "keepe out che vor'ye, or ice try whither your Costard, or my Ballow be the harde; chill be plaine with you". Edward Phillips describes it as "a Word us'd for I in the Western Parts of England" as late as 1706. In his book on the dialect of the West Country in 1869, James Jennings even records the word utchy being used for I in the south of Somerset, which he believes to be a corruption of the disyllabic iche. Utchill for "I will", as well as utchy, was heard by Prince L. L. Bonaparte in the same region in 1875.
When the northern forms of the pronoun established themselves, squeezing ich into the obscure hinterlands of Somerset, capital I most of all became predominant. But why this irregular form appeared at all is somewhat of a mystery. Capital I was first used in the late 14th century, and co-existed with lowercase i and y until the middle of the 15th century, with the i being often dotless. The dot on the minuscule i started to appear as a diacritic in Latin manuscripts of the 11th century to particularize a minim stroke as being a separate letter, especially in the case of -ii where it might be confused with -u. Now the dot is taken to be an inherent part of the glyph.
One of the theories about the capital is that it represents the same kind of particularization, but the dot was already becoming available, and moreover even the absorption of the pronoun into surrounding words (e.g. icham, ichill, ichot) doesn't put it in an ambiguous context. As well as the co-existence problem, any theory of the capitalisation that somehow depends on i being too typographically insignificant or mistakable has to contend with the fact that John Wyclif's Bible from 1388 uses uppercase Y for the pronoun in a medial setting. Also relevant may be the fact that John Barbour's 1375 poem The Bruce (the first purely Scottish literary work), whose texts are from the 15th century, uses both Ic and Ik capitalised.
Charles Bigelow proposed in 1998 that the change may have been phonological in nature, reflecting the diphthongisation that occured changing ɪ to əɪ (I to @I in X-SAMPA). The original pronunciation is apparent in the enclitic absorption of the pronoun as a suffix in words such as haddy and hauy, i.e. I had and I have, both of which were used by Chaucer in his Treatise on the Astrolabe. But if Barbour's Ic and Ik have to be taken into account, this theory is made unlikely. Moreover, it would seem more natural to indicate diphthongisation by using an actual diphthong, as happens sometimes with Scots such as in "Aw was up at Allokirk the day, an' div ye ken what the craiturs war sayin'?" from S.R. Crockett's Stickit Minister of 1893, though the diphthong is much less emphatic in Scots than English.
Another theory is that of ego. Though it is without direct precedent, some languages do capitalise the second person pronoun. In German, Sie and its declensions are capitalised, as are Lei and Loro in Italian, and occasionally U in Dutch. These are all the formal versions of the second person pronoun. In Polish, pronouns referring to a recipient are always capitalised, irregardless of grammatical person. These are all the opposite of ego, of course, intended to flatter or show respect to the party being communicated with; it'd be an interesting psychological idiom if English did the inverse, but the fact that it lost its informal second person pronoun somewhat negates that. The theory of ego falls at the fact that the idiom of capitalisation didn't take hold on any of the other forms of the pronoun, even though as with Y, Ik, and Ic it did appear rarely.
Though they don't stand up to inspection when taken individually, a combination of the theories could be the answer. Some intangible mixture of aesthetics, convention, phonology, clarity, egotism, and a very loose orthography seem to have conspired in the period of English's greatest upheaval to produce a form which appears natural today but makes as much etymological sense as the l in could or the b in debt. Though we crave discrete atomic answers, language is driven by consensus and convention. Even though there is no historical rationale for the capitalisation, we're forced to retain it because of convention. But as with all conventions, they can change rapidly and unpredictably, and such seems to have been the case with the first person singular pronoun.
Cite: Palmer, S.B. (2005). "Origins of a Pronoun", in: What Planet is This?
Archival URI: http://inamidst.com/notes/pronoun