Musings on Future English, aka Cumberland Scratchpad

[Moose-tache]

I love two things: future English conlangs, and making everyone else listen to my inane opinions. I've been working on a future dialect of American English, and it's been an eye opening experience. Something I want to talk about today is vowel perception.

It turns out, the various NA dialects are often distinguished by pretty subtle vowel changes. What's more, they are often very naturalistic changes that could easily spread or even happen spontaneously in other places. At the same time, people have pretty clear ideas about dialects; they recognize patterns and asign them to social categories quite readily. So, for example, if I move the Canadian TRAP vowel up and to the left slightly, and turn the phonetic nasalization knob a couple of ticks, I will trip people's "hillbilly detector," and they will ask why I made Canadian English into Southern English, even though these changes are unrelated to Southern English. Likewise, it's perfectly natural that a future dialect might spontaneously merge the PALM and LOT vowels at some point, at which point people will ask why I made everybody a Californian. Unless you're striking out into weird-for-the-sake-of-weird territory with phonemic tone and pharyngeals, you're going to be playing a game of Twister crowded with countless other similar dialects. This isn't a problem, necessarily, but I anticipate it will be an issue when presenting the language to readers.

Has anyone worked on a future English conlang? Did you find the vowel space to be treacherous and territorial?

[Travis B.]

I've tried working on a Future English, and it's hard to not make it an extension of the English I speak as, well, it's the English I'm most familiar with the phonetic details of, and it's a weirdass variety to begin with (seriously it has lots of stuff going on). Of course, going from this starting point makes it easy to create something truly unintelligible with modern-day English that has some thin basis in reality. The only other choice for me really is to take General American and just make random changes to it, as General American is the main other English variety I hear in Real Life (what I hear in foreign murder mysteries aside); however, the random changes part is because I do not have much to work off of that is interesting about it.

[zompist]

So, for example, if I move the Canadian TRAP vowel up and to the left slightly, and turn the phonetic nasalization knob a couple of ticks, I will trip people's "hillbilly detector," and they will ask why I made Canadian English into Southern English

This sounds like what happens to A.Z. Foreman (@azforeman on Twitter), who does highly rigorous readings in medieval languages, including English, French, Portuguese, Chinese, Hebrew, etc. Invariably people say that his 17th century English "sounds like Irish" or whatever.

I think you shouldn't worry about that sort of reaction. People will hear an unfamiliar accent in terms of their own filters, especially if they can more or less understand the text. It doesn't mean that their reaction is an accurate description.

[Moose-tache]

I've tried working on a Future English, and it's hard to not make it an extension of the English I speak as, well, it's the English I'm most familiar with the phonetic details of, and it's a weirdass variety to begin with (seriously it has lots of stuff going on). Of course, going from this starting point makes it easy to create something truly unintelligible with modern-day English that has some thin basis in reality. The only other choice for me really is to take General American and just make random changes to it, as General American is the main other English variety I hear in Real Life (what I hear in foreign murder mysteries aside); however, the random changes part is because I do not have much to work off of that is interesting about it.

Oh, please do make an extension of your idiolect. I'm working from a conservative NA dialect similar to GA, but I would love to hear some Newfies from the year 3000, or maybe future West Pennsylvanian, or St Bernard Parish in space.

[Moose-tache]

OK, I guess this is a scratchpad now. Feel free to ignore my rantings and ravings.

One of the goals I had with my future English was to improve on Justin B Rye’s Futurese. Apparently it’s pretty common to be inspired in this way, as Jan Misali made a video on this topic. Rye’s Futurese is based on some plausible and fascinating ideas, but executed in a way that is a little too neat and tidy, and besides the whole thing is poorly documented. It comes across as more of a what-if question than a conlang. Because I was inspired by Futurese, I started with a sort of conservative GA (which I actually speak and which does exist in large swaths of the country (I also worked out the interactions between GA, AAVE, Chicano, Canadian, and other dialects, which seems to be pretty rare in these types of projects(also while I’m on the topic, why do future conlangers treat North American Spanish as if it’s one dialect when they’re making loan words?))). I like the idea of vowel breaking that forms the backbone of Rye’s system, but I wanted to make it messier and more irregular, while still being phonemic and systematic. Below I will attempt to explain what happens to stressed vowels in Cumberland, a name I just made up on the spot for my future English conlang in the first couple of centuries of the third millennium.

First, lax vowels break before voiced consonants in the same word. The outcome of DRESS and TRAP, and the outcomes of PALM and CLOTH are the same. This creates the falling diphthongs ɪə, ɛə, ʊə, and ɔə. Some complications apply. Previous to this, nasal vowels were formalized, and although they are never phonemic across the entire history of Cumberland, they are affected differently by sound changes and pair with different oral vowels in different environments. In this case, the nasal is ignored for breaking before another consonant. This means that “can” experiences breaking, but “can’t” does not. The KIT vowel also merges with COMMA in certain scenarios which does not experience breaking. Meanwhile, AAVE universally merged FOOT and COMMA, which causes some loans into Cumberland to behave in weird ways. Oh, and did I mention that flapped d does not cause breaking? Apparently it was analyzed as flapped t. A similar process applies to some instances of g, suggesting that in some subdialects there must have been a parallel flapping applied to the velar consonants, now long gone.

Soon after this there is a process of vowel clipping before voiceless consonants (including, as mentioned, flapped d and coda nt). Once again DRESS and TRAP have the same realization, as do PALM and CLOTH. Since these sounds merge as schwa in unstressed syllables, this constitutes a huge vowel merger. Clipping applies to the lax vowels, but also the FLEECE and GOOSE vowels.

These two sound changes create two nearly-symmetrical systems of vowels. In the high front vowels, FLEECE appears word-finally or before voiced consonants. KIT appears before voiceless consonants, and KID appears before voiced consonants. In mid front vowels, FACE appears in all types of syllables, while TRAP is only before voiceless consonants and TREAD is only before voiced consonants. This arrangement is mostly matched by the back vowels, with some exceptions. While TREAD cannot appear word-finally, its counterpart PAWN can. Also, other than the borrowings from AAVE, there is no corresponding KIT-COMMA split for the FOOT vowel. This system first gains phonemic load when clusters undergo regressive voicing assimilation. For example, “Daphne” is pronounced with a clipped TRAP vowel and a voiced labial fricative. Some other changes to coda consonants increase the number of minimal pairs (still no categorical loss of voicing). But overall the phonemic distinction between broken and clipped vowels is minor.

Rhotic vowels follow their own logic, as do, to some extend, lateral vowels, which inhabit a liminal space between vowels and vowel-consonant sequences.

[Travis B.]

If I am going to start off with the English here for a Future English, I'd have the following initial changes (mind you that these aren't "future" at all):

• Systematic vowel length allophony conditioned by historical fortisness/lenisness of following obstruents (long by default)
• Coda (pre)glottalization of fortis plosives
• Coda devoicing of lenis obstruents
• Optional devoicing of initial lenis obstruents (note that voicing is more likely to be preserved, except for /dʒ/, in careful speech)
• Full NCVS except that /æ/ only breaks before nasals and /ɑ/ has a back realization adjacent to /r w h kw gw/
• Canadian Raising (including of /ɑr/) but with a back rather than front starting point for /aʊ/
• Mid endpoints for /aɪ aʊ/
• Aggressive l-vocalization producing unrounded vowels and semivowels except when adjacent to a rounded vowel or semivowel, where then the result is rounded.
• Massive intervocalic elision of /t d n nt nd/ and sporadically /b v ð/ and even in a few cases /l/; note that /nd/ sometimes becomes [n] or even [nː] instead
• Elision of nasals before fortis obstruents; this leaves the preceding vowel strongly nasalized
• Otherwise weak nasalization of vowels before nasal consonants, even if they were elided
• Creation of new diphthongs from lax vowels before elided intervocalic consonants, unless /ər əl əm ən/ followed, where then the schwa was lost and the lax vowel was lengthened (so it would be long if it was short previously, and it would be overlong if it was long previously)
• Realization of /r/ normally as a pharyngealized uvular approximant (a "bunched r") except initially, where it is also labialized.
• Affrication of /t/ and /d/ before /r/, which is then realized as a postalveolar-uvular coarticulated approximant without pharyngealization. Note that in this case /s/ preceding the plosive is normally alveolopalal.
• Realization of /eɪ oʊ/ as [e̞] and [o̞]~[ɵ̞] except /oʊ/ may be pronounced as [o̞u̯] or, when unstressed, [əu̯] before a vowel or finally (especially when stressed, e.g. in hello)
• Realization of /dʒ/ as voiceless in all cases
• Palatalization of alveolar and postalveolar consonants before /w u ʊ ər/ to palatalized alveolars and alveolopalatals respectively
• Palatalization of sibilants before /p b t d tʃ dʒ m n r l w/ and after /k g r l/ to palalized alveolars and alveolopalatals
• Frequent realization of palatalized /t/ as an alveolopalatal affricate; some people also do this to palatalized /d/
• Spreading of palatalization to all adjacent consonants that can be palatalized
• Centralization of /u ʊ oʊ/ after and between non-dental coronals
• Assimilation of /t d/ to following obstruents, geminating them (note that vowel length and (pre)glottalization is preserved)
• Assimilation of /bm bəm/ and /dn dən/ to [mː] and [nː] without nasalizing the preceding vowel.
• Frequent alveolar or dental stopping of initial /ð/ except when following /n/ or a sibilant, where then it tends to assimilate with it instead (producing a dental nasal with /n/ instead)
• Frequent realization of /θ/ as a dental stop or affricate
• Lowered pitch with lenis obstruents in syllable onsets; where new syllables are produced via elision of a fortis obstruent, a rising pitch arises.

As you can see, a load of stuff is going on to work with. Of course, you're probably wondering "how in hell does Travis B. have such a batshit dialect?", which you probably have been wondering for years at this point. Frankly, though, I suspect at least half of this stuff is also going on in many other NAE varieties, but rather people simply do not notice it due to being accustomed to it without it actually being mentioned (hence why people who have heard my voice have commented that I don't seem to have that odd of a lect).

[Moose-tache]

It's that time again, when I pick people's brains about future English.

I guess it's worth discussing whether we count literary devices like Nadsat, mostly made as window dressing by non-linguists. Some of them are pretty effectively used (like David Masson's in The Transfinite Choice, a weird short story set in an overcrowded future Earth; pretty dull today, but in 1966 this kind of Malthusian panic was slightly ahead of its time), but they are rarely well thought-out as languages. The sound changes tend to be unsystematic, and the grammar is almost always identical to the working class dialects of contemporary English in the author's home country.

Fully fleshed-out conlangs are also out there, but many of them are poorly documented. A simple googling will reveal countless threads form the last twenty years where someone says "I'm making a future English; how should I do comparative adjectives?", followed by silence. Some of them are reasonably well documented. For example, Trigedasleng has a dictionary and a fair bit of grammatical information floating around. Like all conlangs, it was created by David Peterson, who usually does a good job of making his creations available to the public in an easily digestible format. But Trigadeslang is deliberately a little bit... dumb. It simplifies some aspects of English grammar and phonology in a very blunt way. This may be because in-universe it was invented by some rando, or it may be because actors on the CW cannot be bothered to learn a real language. Better designed conlangs tend to be worse when it comes to documentation. Justin Rye's Futurese has no grammar or dictionary at all. Jan Misali complained about Futurese, and made a response, Megalopolian, that also has no real documentation.

It seems that "future English" is a pretty broad, popular category of conlangs that nevertheless has no Ur-chetype. Fantasy artlangs have Quenya; Alt-history bogolangs have Brithenig; Sci-fi fanlangs have Klingon. IALs have (for better or worse) Esperanto. Most types of conlang have some kind of granddaddy looming over the upstarts. But it seems that no one has ever made a complete future English that we can look at that doesn't suck. What do you all think about this type of language? Do you have any opinions about how they should be approached? Do you have a favorite? Am I wasting my time?

[also, I just now invented the word "ur-chetype," and I am extremely proud of myself.]

[Travis B.]

If I had to make one con-change starting from my batshit variety, I would make it be to make elision and assimilation obligatory in many common words, phonologizing things such as vowel length (with unspecified, short, long, and overlong vowels!), consonant length, secondary diphthongs, phonemic vowel nasality, and even adding pitch accent derived from previously allophonic pitch variation conditioned by fortisness versus lenisness. This alone, without any other changes, dramatically changes the underlying phonology despite being essentially identical to how I already speak; the only real surface variation is that forms to which these changes are not applied would disappear, except perhaps in markedly conservative speech. This raises important questions, though, such as do underlying phonologies actually matter, if the surface forms are nearly identical?

[Emily]

It's that time again, when I pick people's brains about future English.

I guess it's worth discussing whether we count literary devices like Nadsat, mostly made as window dressing by non-linguists. Some of them are pretty effectively used (like David Masson's in The Transfinite Choice, a weird short story set in an overcrowded future Earth; pretty dull today, but in 1966 this kind of Malthusian panic was slightly ahead of its time), but they are rarely well thought-out as languages. The sound changes tend to be unsystematic, and the grammar is almost always identical to the working class dialects of contemporary English in the author's home country.

Fully fleshed-out conlangs are also out there, but many of them are poorly documented. A simple googling will reveal countless threads form the last twenty years where someone says "I'm making a future English; how should I do comparative adjectives?", followed by silence. Some of them are reasonably well documented. For example, Trigedasleng has a dictionary and a fair bit of grammatical information floating around. Like all conlangs, it was created by David Peterson, who usually does a good job of making his creations available to the public in an easily digestible format. But Trigadeslang is deliberately a little bit... dumb. It simplifies some aspects of English grammar and phonology in a very blunt way. This may be because in-universe it was invented by some rando, or it may be because actors on the CW cannot be bothered to learn a real language. Better designed conlangs tend to be worse when it comes to documentation. Justin Rye's Futurese has no grammar or dictionary at all. Jan Misali complained about Futurese, and made a response, Megalopolian, that also has no real documentation.

It seems that "future English" is a pretty broad, popular category of conlangs that nevertheless has no Ur-chetype. Fantasy artlangs have Quenya; Alt-history bogolangs have Brithenig; Sci-fi fanlangs have Klingon. IALs have (for better or worse) Esperanto. Most types of conlang have some kind of granddaddy looming over the upstarts. But it seems that no one has ever made a complete future English that we can look at that doesn't suck. What do you all think about this type of language? Do you have any opinions about how they should be approached? Do you have a favorite? Am I wasting my time?

[also, I just now invented the word "ur-chetype," and I am extremely proud of myself.]

i personally feel that all future english conlangs should be derived from the way that i, emily, speak right now (socal gay, descended from the famed valley girl accent of yesteryear)

[Travis B.]

i personally feel that all future english conlangs should be derived from the way that i, emily, speak right now (socal gay, descended from the famed valley girl accent of yesteryear)

How interesting linguistically though is socal gay-speak as a starting point though?

[Ares Land]

It seems that "future English" is a pretty broad, popular category of conlangs that nevertheless has no Ur-chetype. Fantasy artlangs have Quenya; Alt-history bogolangs have Brithenig; Sci-fi fanlangs have Klingon. IALs have (for better or worse) Esperanto. Most types of conlang have some kind of granddaddy looming over the upstarts. But it seems that no one has ever made a complete future English that we can look at that doesn't suck. What do you all think about this type of language? Do you have any opinions about how they should be approached? Do you have a favorite? Am I wasting my time?

[also, I just now invented the word "ur-chetype," and I am extremely proud of myself.]

A scenario I'd love to explore sometimes, anyway, is to start from 'Globish' (for lack of a better word) -- English being adopted by non native speakers and essentially replacing various non-native languages.

It'd be a very difficult project. There are numerous variants of English, and it's hard to predict which would be used as a global model in a century or so. Then there is the unresolved question of how language would evolve under the influence of mass media -- plus the fact that older variants of the language would available basically forever, and to most everyone.

[Travis B.]

A scenario I'd love to explore sometimes, anyway, is to start from 'Globish' (for lack of a better word) -- English being adopted by non native speakers and essentially replacing various non-native languages.

It'd be a very difficult project. There are numerous variants of English, and it's hard to predict which would be used as a global model in a century or so. Then there is the unresolved question of how language would evolve under the influence of mass media -- plus the fact that older variants of the language would available basically forever, and to most everyone.

How much influence does media have on native speakers, though? (It definitely has influence on non-native speakers, it should be noted.) I have heard plenty of media in not just NAE but also EngE, but EngE has not influenced my speech one bit. The main influences on my speech have been people in Real Life around me, not media.

Of course, a bit question is whether the future of English is diglossia or standard varieties stomping out dialects in areas or something in between. We have seen much stomping out of local dialects in various places, but in many cases those local dialects have been replaced by regional dialects rather than by standard varieties. But even then will said regional dialects stay under the influence of standard varieties so as to not differ that much from them, or will diglossia between the two arise?

[Ares Land]

All good questions.
My money was on media not influencing native speakers much or at all but I read in this very place about people being influenced by American English so it's not as clear cut as that.

[Travis B.]

All good questions.
My money was on media not influencing native speakers much or at all but I read in this very place about people being influenced by American English so it's not as clear cut as that.

Overall, though, most of the non-NAE-speaking native English speakers here seem to be firmly non-NAE phonology-wise (even though I think I once recall someone, an Australian IIRC, commenting on that they'd picked up rhoticity from NAE media). This is what I would expect, considering that while standard varieties of NAE and of EngE and AusE are essentially fully crossintelligible (even though my parents would deny this, considering that they watch British TV with subtitles, which I find to be silly since I can understand it perfectly well without them), the underlying phonological systems are considerably different — you cannot just turn a non-rhotic variety of EngE or a variety of AusE into a rhotic variety because for speakers of such varieties the only indication of where rhotics "should" be is orthography.

[Emily]

i personally feel that all future english conlangs should be derived from the way that i, emily, speak right now (socal gay, descended from the famed valley girl accent of yesteryear)

How interesting linguistically though is socal gay-speak as a starting point though?

well afaict most other dialects don't have a diphthong in "what" lol