All right sisyphus, now I'm convinced I need to revise it.
Many of your questions can be answered by a good grammar that has notes on historical linguistics (e.g. A&G), but I'll answer with what I know personally. There is a particular lack in the article of defining which formations happen in which conjugations, an important and useful thing to know.
sisyphus wrote:
Indeed, fascinating. i actually find this kind of thing at least as interesting as learning the language itself. Is that just sad, or some kind of sickness? Maybe it's my over-systemising brain finding something it can finally get its claws into, rather than leaving it to the vagaries of backing store.
If it's a sickness, you aren't the only one here to have it.
to form the perfect stem we often (but not always) add an S to the present stem
Cur, Benissime, 's' adaugemus? "Because i say so" may be a valid answer, but unhelpful.
The reason is that the Latin perfect system absorbed the aorist tense/aspect familiar to those who have studied Greek. The Latin aorist was formed, as in Greek, by adding an S/sigma to the verb root, so the formation is preserved in the Latin perfect (alongside the other perfect stem formations). Examples of perfect stems ending in S are abundant in Latin, mostly of the third conjugation, and these are all using a formation of aorist origin. However, it is not just the formations that joined - the Latin perfect tense has absorbed much of the meaning of the aorist as well, which explains why the Latin perfect is much different from the English or Greek perfect. It is much more aspectual and not frequently used in its "true perfect" sense.
the much feared "agtum"
But why is/was it considered more awkward to pronounce "agtum" than the much feared "mittere"? Is this due to the education of my palate?
It's hard to say why exactly, but in many languages for the sake of fluidity, a voiced mute consonant before a sibilant or unvoiced lingual becomes unvoiced. In other words: Latin B and G followed by S or T tend to turn into P and C respectively (D->T doesn't seem to happen for some reason). This is for the sake of easier and faster pronunciation, and you can see examples of the same process in English. There are other situations where unvoiced consonants become voiced, but that is more theoretical in Latin since it wasn't represented in spelling. You can certainly see that process of "unvoicing" in English though where the voiced -(e)d of the past tense is pronounced and sometimes written -t when preceded by certain sounds (compare the pronunciations of -ed in spilled and asked, or the very different spellings but similar sounding endings in learnt and risked, where the pronunciation of the d/t consonant varies based on what letter precedes it).
English has more mute consonants than Latin, so the changes that tend to take place are: B->P, G->K/C, V->F, Z->S, etc. There are also other relationships, such as those between the letter B, P, F, and V; an aspirated P (ph) is an unvoiced form of B, but PH tends to turn into F, and V is the voiced form of F, so there is a continuum, [ B (BH) <-> P <-> PH <-> F <-> V ], in which a consonant may be standing in place of another consonant that would at first seem totally dissimilar to and unrelated to it.. This particular rule singlehandedly describes why English has
brother, Latin has
frater, and Greek has
phrater.
Why does the V in "leave" turn into F in "left" (phonetically lev+t = lef+t), or why is the past tense of "wend" "went"? There are simply rules in languages, including Latin, that govern which sounds can change to which other sounds and when they can do so, and they are amazingly predictable.
The article doesn't mention (more or less) conscious perversions (or cultural adomptions) to prevent confusing clashes with other word forms. i don't know of any verb examples, but the adoption of "deabus" to prevent confusion between the abl. pl. of "deus" and "dea" seems a good illustation (sim. "filiabus").
Can some instances of "suppletion" be explained as similarly being introduced to avoid confusion with other word forms? (And what actually is the history of the use of "went" with "go"?)
I am afraid I don't know of any specific instances of conscious changes being made to perfect tense forms to prevent confusion with other words. Your idea that suppletion could be indicative of this is very interesting, but there aren't many examples of suppletion in Latin to draw on (perhaps fortunately so).
The suppletion of
sum, esse with
fu- is very old and I doubt that they merged for any other reason than their similar meanings (ES "to be", BHEU "to grow" - again bheu->fu- = BH->PH->F).
The suppletion of
fero, ferre and
tul- is, I believe, a borrowing (or stealing) from
tollo.
tollo then, needing a perfect stem, you could say took it from a compound of
fero to avoid confusion with
fero itself. Then you still have confusion with
tollo and
suffero, so I don't know if that really helps explain anything.
I can't think of any other examples of Latin suppletion off the top of my head.
As for "went", AHD can describe it much better than I:
http://www.bartleby.com/61/29/W0092900.html wrote:
ETYMOLOGY: Middle English, from Old English wende, past tense and past participle of wendan, to go.
WORD HISTORY: Why do we say went and not goed? Go has always had an unusual past tense, formed from a completely different root from its present tense. The replacement within a series of inflected forms of one form by a completely unrelated form is called suppletion. (Another, even more extreme, example of suppletion in English is found in the paradigm be, am, are, was, whose forms are originally from four different verbal roots.) The past tense of go in Old English was eode, formed from an unrelated root that has no other verb forms in English. Its modern replacement, went, derives from old forms of the modern verb wend. In Middle English the original past tense and past participle of wenden, “to go, turn,” were wended and wend, respectively. The forms wente and went appeared around 1200 and gradually displaced the older wended and wend. The new past tense wente also took on a new use as the past tense of go, replacing ode. By the beginning of the Modern English period, around 1500, went was no longer used in any other way and was therefore felt to be the normal past tense of go; at the same time, wend acquired the new form wended for its past tense and past participle, meaning “turned.”
H+S=X
This would be more acceptable to me if 'h' had a less aspirated sound than i thought it had in Latin. i don't know how to describe this phonetically, but the implication is that 'h' is pronounced with a slight catch in the back of the throat. Perhaps more the way 'j' is pronounced in Spanish ("Jose")?
That's what I assumed at first but William is correct. Latin H and English H stand for quite different sounds from a historical standpoint. There is a really interesting table in A&G §19:
For some example of how different in derivation the Latin H is from English H, Latin
haedus is cognate with English "goat" (with HaeDus corresponding to GoaT),
hortus with "yard" (HoRTus - YaRD). Conversely, the sound that came into Latin as C tends to come into English as H, as in
CoRD- and "HeaRT",
CORNu and "HORN". This is not to say that by the Classical period the Latin H still had this sound; on the contrary, it was probably even softer than the English H and by at least some people not even pronounced (as demonstrated in Latin's derivative languages).
It's also worth noting that V stands for an originally harsher sound as well, by which
VIV- (vivus) is related to "QUICK",
VENio to "COMe".
Article states that "a" may weaken to "i", and then gives as an example the third parts of "ago, agere, egi, actum" and "facio, facere, feci, factum". i can understand in choral terms the progression through "a", "e" and "i": i presume it's similar?
I don't think the a->i comment belongs there, since it usually only happens in the present tense when words are compounded and creates unnecessary confusion. I can't think of any case in which A changes to I in the perfect tense. The A->E lengthening is relatively rare, occurring only in the 3rd conjugation in about 5 verbs, but those just happen to be some of the verbs you learn first (
facio, capio, iacio, ago and
frango).