Reconstructing ancient US English

[zompist]

From the Paleo-European Languages thread:

Imagine most written sources in English are lost, and linguists in the 7th millenia trying to make sense of US toponymy. Some place names they'd figure out, like 'Long Island', or even 'Brooklyn' and 'Staten Island'. But they'd never find anything satisfying for 'Connecticut', or 'Chicago', or 'New Mexico'.

This sounds like an interesting challenge. Let's say we have

* The names of the US states and their capitals
* The names of its largest rivers
* The names of the fifty largest cities

Assume we know what they are (states, cities, and rivers) but have no dictionaries and no 'modern' cognates.

What could you learn? Can you guess any meanings? Would any substrates be evident?

[Raphael]

I guess it should be possible to figure out "North" and "South" that way, but I don't know about anything else.

[Ares Land]

You mean, besides the obvious Baltic substrate?

The first thing we pick up is 'River'. (I'm assuming our observer gets the name as listed in Wikipedia.)
'Rio Grande' is a weird outlier that is best ignored for now. (Of course, another observer with more data from the rest of the Americas would figure it out and peg it as cognate with River.)

North and South are figured out from the Dakotas and Carolinas.

Maybe we could reconstruct 'West', from West Virginia.

Other prefixes: New, San, Saint and Santa. (the last one being a weird outlier).

'City' can be figured out from Oklahoma City, New York City, Salt Lake City and Kansas City.
We can even figure out from Kansas City that it derives from river, not name.

The meaning of 'City' is obviously, 'a place on a body of water' and I can confidently reconstruct 'Salt Lake' for the 'Great Salt Lake' and 'New York River' for the Hudson River.

I don't know what to make of 'New'. I don't know if being aware of 'Mexico' (the country) would help or not, so let's say that name has been lost to history.

City names tend to be prefixed by San with the variants Santa, Saint; Saint is used in one river name.
Santa might be a variant of San, or just a weird outlier, not related at all. But I think it might be an intermediate form. There seems to be a geographical pattern with San, Saint. Saint being found in the East, San in the West, so perhaps the isogloss runs through Santa Fe.

In any case, we don't know what Saint, San, Sant- may mean, so in the meanwhile we'll just say it's an honorific.

We may also notice a common root: Columb-.
Of course, we see the suffix -on, as found in Yukon, Oregon, Washington, Cimarron, Trenton, Carson City, Jefferson City, Jackson, Boston, Baton Rouge, Houston, Arlington
It seems to have three allomorphs: -ton after s or a nasal, -son after s, r (possibly assimilated in Cimarron).
Other common suffixes: the very common -a and -ia, and the suffix -as, found in Texas, Arkansas and Texas.

[Creyeditor]

Maybe the San/Santa, Rio/River alternations were gender agreement and both were originally adjectival modifiers?

[Ares Land]

Oh, that's an interesting idea.

There's also the fact that both hapaxes Rio and Santa are found in New Mexico. Maybe an hint that a divergent dialect was spoken there.

It's also the one area where 'x' and 'qu' are found !

[dɮ the phoneme]

There's definitely evidence for a suffix -land, from Portland, Oakland, Maryland, Road Island, and Cumberland River, with a possible variant -leans in New Orleans. It might mean "inlet" on the basis of Oakland, Maryland, Road Island, and New Orleans, though this leaves Portland and the Cumberland River unexplained.

[zompist]

Of course, we see the suffix -on, as found in Yukon, Oregon, Washington, Cimarron, Trenton, Carson City, Jefferson City, Jackson, Boston, Baton Rouge, Houston, Arlington

I like this one a lot... a perfect mixture of actual and coincidental resemblances.

What's particularly neat is the reconstructor might well correctly guess that -ton means a place of residence, without knowing that for most of these names the process was place name > surname > place name.

[circeus]

Of course, we see the suffix -on, as found in Yukon, Oregon, Washington, Cimarron, Trenton, Carson City, Jefferson City, Jackson, Boston, Baton Rouge, Houston, Arlington

I like this one a lot... a perfect mixture of actual and coincidental resemblances.

What's particularly neat is the reconstructor might well correctly guess that -ton means a place of residence, without knowing that for most of these names the process was place name > surname > place name.

There would be huge debate regarding whether -ton ought to be regarded as a variant of -town.

[Yalensky]

There would be huge debate regarding whether -ton ought to be regarded as a variant of -town.

Perhaps the -tin in Austin as well. Also note the variations: -son/-sin (as in Wisconsin) and -ton/-tin. And what of -ta of Atlanta and Augusta? Surely the vowel marks gender.

[zompist]

The corpus doesn't include "town".

Some more interesting things:
* -polis is not uncommon: Indiana-, Ann-, and Minne-
* -ville twice: Nashville, Louisville
* Fort in Frankfort and Fort Worth; tempting to look also at Hartford
* Spring in Springfield, Colorado Springs
* -son is pretty common: Jackson, Jefferson, Carson, Madison, Tucson, Jacksonville.

It's tempting to expand the database to the top 300 cities. We'd certainly get a lot more morphemes.

Since this is modeled on investigating early European toponyms, it seems fair to say we get the actual locations too. Then we might, say, note that "valley" names appear in valleys.

[bradrn]

The river names show some interesting patterns. One prominent element is *mis-, in Mississipi and Missouri. Des Moines and Tennessee might have a variant of that (*mins- and *-nes- respectively). The same area also has Ouachita, Wabash, White, all starting with *wa(i)-. Notably, both of these elements are restricted to an area centred around modern-day Memphis… could there have been a substrate in this area?

[dɮ the phoneme]

How well can we actually read the script? I'm assuming these names are attested solely orthographically (on ancient "welcome to ____" signs, perhaps? ), and we've reconstructed which areas they refer to on the basis of archeological evidence ("Welcome to Texas" signs only being found in Texas). But, presumably, we can't actually read them, if no Ancient English dictionary survives. Do we just have general knowledge of the latin script, or are the phonetic values entirely unknown as well?

[bradrn]

How well can we actually read the script? I'm assuming these names are attested solely orthographically (on ancient "welcome to ____" signs, perhaps? ), and we've reconstructed which areas they refer to on the basis of archeological evidence ("Welcome to Texas" signs only being found in Texas). But, presumably, we can't actually read them, if no Ancient English dictionary survives. Do we just have general knowledge of the latin script, or are the phonetic values entirely unknown as well?

Given that this is ‘modeled on investigating early European toponyms’, as zompist said, I would imagine that it’s precisely the opposite: we know how they’re pronounced, but not their orthography.

[Ares Land]

How well can we actually read the script? I'm assuming these names are attested solely orthographically (on ancient "welcome to ____" signs, perhaps? ), and we've reconstructed which areas they refer to on the basis of archeological evidence ("Welcome to Texas" signs only being found in Texas). But, presumably, we can't actually read them, if no Ancient English dictionary survives. Do we just have general knowledge of the latin script, or are the phonetic values entirely unknown as well?

I don't know, I've assumed we magically know both spelling and pronunciation

To further complete the analogy, we should really mangle these names through sound change and poorly attested ancient languages

[2+3 Clusivity]

/*monT/ > Rich-mond, Ver-mont, mont-pelier, mont-ana, mont-gomery .... probably researchers can guess something with hills/mountains with the exception of Montgomery. Richmond seems to be a bit of an outlier but it's one of the fewer items in final place.

/*con/ > con-cord, con-necticut, Wis-con-sin

Building on the form pointed out above, I'd note a broader /*məS ~miS ~ wiS/ mar-yland, mas-sachusetts, mich-igan, mis-sissippi, mis-souri, wis-consin, ?min-nesota, ?win-nipeg ??wyo-min-g. ... Maybe something to do with coastland, rivers, deltas. Interestingly, this form itself seems to be followed by /*-s(V)-/ or /*-ni- ~ -ne-/ frequently.

[dhok]

A cognate to miss- also exists in Massachusetts, though that might not be picked up on. (There's a suburb of Boston called Wuchusett, but it's probably not big enough to qualify in the data).

Manhattan looks like it has the -tən morpheme. Probably the Man- would be taken as cognate to the mi- of Minnesota and Missouri--maybe Massachusetts as it's much closer. (All of these connections are incorrect--Manhattan is from PA *menahanwi.

Connecticut has the Algonquian locative, as does Massachusetts, though the latter's -s might throw things off; there's also Nantucket. (If rivers survive, islands probably do too). The locative in -tt is only found in New England, and probably won't be reconstructible, but one in -k may be, though maybe incorrectly. (I'm going to assume we have the top 300 cities to work with). In addition to correctly-identified placenames with a -k locative like Merrimack, Rappahannock, Potomac, Chesapeake, Kennebec, Housatonic...in fact, a reasonable guess would be that the -k locative means 'river', like European rivers in R-; Newark might be taken as a rhotic hypercorrection. Other incorrectly-identified k-locatives might include Norfolk and Lubbock. The original nasal of the locative (PA -enki) doesn't survive very well, though if you include Canada then you have -ing names. Lansing might be misidentified on that basis.

-chester is likely to be recognizable: Manchester (NH) and two Rochesters (NY and MN) possess it. Not so clear is whether it would be connected to Lancaster (CA) and the particularly impenetrable [wʊstə(r)], Massachusetts.

[sasasha]

Linguistic quackery of the 7th millennium:

-as, -is, -es and -os can teach us much about the ancient world.

-as only is found in state names.
-es, -is, and -as are all found in city names.
-os is found only in river names.

I believe -Vs to have been a locative suffix showing agreement with the grammatical gender of suppressed head nouns - or perhaps an animacy system. Why, yes - rivers were always conceived of as highly animate by ancient peoples, so -os represents the highest animacy class! Or just as plausibly, the ancient noun classes were based on size: -as was most likely applied to areas large enough to constitute 'the whole known world' of ancient peoples.

Ancient Idahoan was clearly a conservative dialect, preserving the older form -ise in 'Boise'.

In some dialects it seems the locative -s was lost, yielding forms ending in the thematic noun-class vowels. 'Maine' may be such an example - perhaps as a small state, it did not merit inclusion in the set represented by 'Texas', 'Kansas' and 'Arkansas'. ('Arkansas' and 'New York' seem to contain the same (C)Vrk morpheme.)

Due to the rather puzzling distribution of "full -s" and "-s-drop" forms, it is likely that the locative -s was preserved only by local orthographic convention. Sociolinguistic factors were clearly wreaking havoc with the twenty-first century phonology and orthography and a general lack of social cohesion is evident, without a strong centralised administration to keep the language in check - typical of a primitive ancient society.

[bradrn]

Linguistic quackery of the 7th millennium:

-as, -is, -es and -os can teach us much about the ancient world.

-as only is found in state names.
-es, -is, and -as are all found in city names.
-os is found only in river names.

I believe -Vs to have been a locative suffix showing agreement with the grammatical gender of suppressed head nouns - or perhaps an animacy system. Why, yes - rivers were always conceived of as highly animate by ancient peoples, so -os represents the highest animacy class! Or just as plausibly, the ancient noun classes were based on size: -as was most likely applied to areas large enough to constitute 'the whole known world' of ancient peoples.

Furthermore, the *-as element is restricted to a small set of nearly-contiguous states, in approximately the same area as the *mis- and *wa(i)- elements I mentioned earlier.

In some dialects it seems the locative -s was lost, yielding forms ending in the thematic noun-class vowels. 'Maine' may be such an example - perhaps as a small state, it did not merit inclusion in the set represented by 'Texas', 'Kansas' and 'Arkansas'. ('Arkansas' and 'New York' seem to contain the same (C)Vrk morpheme.)

Surely /ˈɑɹkənsoː/ would be -s-drop as well?

Due to the rather puzzling distribution of "full -s" and "-s-drop" forms, is likely that the locative -s was preserved only by local orthographic convention. Sociolinguistic factors were clearly wreaking havoc with the twenty-first century phonology and orthography and a general lack of social cohesion is evident, without a strong centralised administration to keep the language in check - typical of a primitive ancient society.

[sasasha]

Linguistic quackery of the 7th millennium:

-as, -is, -es and -os can teach us much about the ancient world.

-as only is found in state names.
-es, -is, and -as are all found in city names.
-os is found only in river names.

I believe -Vs to have been a locative suffix showing agreement with the grammatical gender of suppressed head nouns - or perhaps an animacy system. Why, yes - rivers were always conceived of as highly animate by ancient peoples, so -os represents the highest animacy class! Or just as plausibly, the ancient noun classes were based on size: -as was most likely applied to areas large enough to constitute 'the whole known world' of ancient peoples.

Furthermore, the *-as element is restricted to a small set of nearly-contiguous states, in approximately the same area as the *mis- and *wa(i)- elements I mentioned earlier.

I believe we have uncovered evidence of a mysterious substrate, henceforth known as 'Proto-Miswasian', which I tentatively posit may be identified with the '7-Eleven' material culture known from recent archaeological enquiry.

In some dialects it seems the locative -s was lost, yielding forms ending in the thematic noun-class vowels. 'Maine' may be such an example - perhaps as a small state, it did not merit inclusion in the set represented by 'Texas', 'Kansas' and 'Arkansas'. ('Arkansas' and 'New York' seem to contain the same (C)Vrk morpheme.)

Surely /ˈɑɹkənsoː/ would be -s-drop as well?

Indeed - an intermediate stage! Thus proving the theory.

[bradrn]

Linguistic quackery of the 7th millennium:

-as, -is, -es and -os can teach us much about the ancient world.

-as only is found in state names.
-es, -is, and -as are all found in city names.
-os is found only in river names.

I believe -Vs to have been a locative suffix showing agreement with the grammatical gender of suppressed head nouns - or perhaps an animacy system. Why, yes - rivers were always conceived of as highly animate by ancient peoples, so -os represents the highest animacy class! Or just as plausibly, the ancient noun classes were based on size: -as was most likely applied to areas large enough to constitute 'the whole known world' of ancient peoples.

Furthermore, the *-as element is restricted to a small set of nearly-contiguous states, in approximately the same area as the *mis- and *wa(i)- elements I mentioned earlier.

I believe we have uncovered evidence of a mysterious substrate, henceforth known as 'Proto-Miswasian', which I tentatively posit may be identified with the '7-Eleven' material culture known from recent archaeological enquiry.

Ah, yes — archaeological evidence! Note also the following material cultures:

• The ‘Arches culture’, distinguishable by their unique buildings decorated with three golden arches. Clearly these acted as temples to an as yet unknown god.
• The ‘slab culture’, apparently widespread across North America, with a concentration of small black slabs. The majority of these slabs are sparsely decorated, typically with a stylised apple or face — a devotional symbol, perhaps? But many of these slabs are brightly decorated on one side, clearly showing that they were used predominantly as a drawing surface. Each slab has a glossy side as well, which was evidently useless for painting and so was presumably completely ignored.

Surely these can all be identified with substrates as well.