This is several months belated, but I thought I’d toss this up here: A few final observations on the Conlang Seminar I took at Swarthmore this previous semester.
Firstly, some thoughts on the structure of the class: A class like this really does need to be a seminar – that is, to be seminar-sized. In many ways, the class was attempting to be a “conlanger’s workshop”, with lots of peer feedback. In particular, the professor insisted on having a third of the class present the state of their conlang once per week. With 22 people in the class, this meant that there was never much time for actually giving feedback, or for class discussion; rather problematic. I love the idea, it just needed to be limited to 10-12 people to be practical.
Regarding the content of the class: As I said to the professor several times, I was slightly disappointed that the only established conlangs that we looked at in any detail were Esperanto and Klingon. That seems (to me) to be problematic for two reasons: Firstly, that neither of those are languages created by experienced conlangers, and both show many of the hallmarks of first conlangs; second, that they only cover two points of the Gnolli Triangle. Why no engelang on the list? Yes, we talked briefly about Lojban, based entirely on the discussion in Okrent’s In the Land of Invented Languages, but didn’t look at any primary sources for it. I understand the professor’s argument that Esperanto and Klingon are some of the only conlangs with significant secondary source material (with Tolkein’s work as probably the only third), but I still would have preferred more variety. As I suggested to him: If I were to run this course, at least one weeks assignment would be to pick a conlang from somewhere out in the community, and write secondary source material. Go pick some language that interests you, and review it!
Regarding the products of the class: There were some cool languages to come out of this class! Unfortunately, I don’t have documentation on any of them, and don’t really expect any of the students to post documentation. For me, though, the class was a wonderful chance to observe patterns and trends among new conlangers. Most of the class was made up of first-time conlangers – there were only three of us who had done any conlanging before. The new conlangers, in my mind, fell into two observable categories: Those with significant prior linguistic knowledge (which mostly meant seniors – those who’d had the most time to take other linguistics classes) and those who didn’t (mostly underclassmen). Each of these groups, I felt, could be subdivided into more-or-less two groups based on the pattern their conlanging followed.
The new conlangers with less prior linguistic knowledge mostly went what I call the Kitchen Sink Included route. You know the pattern: The conlanger who just looks around at natlangs, picks all their favorite features, and sticks them all into the language, willy-nilly. The resulting languages feel ad hoc, unfocused, experimental. I certainly went through this stage: All of my early drafts of tükwäi (which are many and varied!) look more-or-less like this. And my experience from watching new conlangers in this class and online is that its a very common pattern. And why not? Conlangs are meant to be personal, right? They’re you’re chance to make the perfect language, just for yourself! It’s a perfectly comprehensible mindset; it also (in my mind) very rarely produces conlanging that anyone else would want to spend much time looking at. (And is it just me, or do nearly all conlangers who start with this move on to create a tri-consonantal root morphology language?)
A few of this group of new conlangers went a different route, conditioned by a very similar mindset. These people, rather than focusing on linguistic features that they like, focused on cultural or world-building aspects that they like, and then built conlangs that matched that culture. I call this the Whorfian Utopia pattern – and to be clear, it’s not mutually exclusive with the above (as evidenced, again, by my own early conlanging!). In our seminar, most of these people seemed to follow the Magical Language trope, with languages that either were magic or reflected magical aspects of the universe that they had created. While I firmly believe that good artlanging involves creating languages that give a sense of cultural background, languages that follow this trope too heavily can feel a bit trite, a bit just-so. And, of course, the utopias that they’re created for are always highly personal, making it hard for the conlang to speak to anyone reading about it.
The most common pattern amongst the students with slightly more linguistic background is one that I like to call All the Cases! You know the archetype: The language that has 503 individually-inflected noun cases, each with increasingly complex Latinate names and super-specific uses. Oh, and they’re all fused with the marking for 7 different grammatical numbers, too. (How many times in the history of the Conlang-L have we had “Name that Case” discussions, basically for this purpose?) Of course, it’s not just cases: Tense, aspect, mood, and phoneme inventory are also commonly subject to the same treatment. We had one (really awesome) language in the seminar that did this with phoneme harmony – it had multiple types of both consonant and vowel harmony, each moving in different directions and subject to different restrictions. (It also had every verb mood or aspect known to human kind, and then some.) This pattern, in my mind, is fundamentally similar Kitchen Sink conlanging, except that the conlanger in question has focused in on one particular favorite feature, and really worked on that. The results are usually less grab-bag then Kitchen Sink languages, but there’s still something fundamentally unnatural about them: When you learn about them, you can see all the evidence, right their on the surface, of the time and effort it took the conlanger to come up with all of those Latinate case names.
All of the above patterns mostly apply to people who were doing more-or-less naturalistic artlangs; a small portion of the more linguistically experienced students in the seminar headed the engelang route. In particular, they were all doing what I think of as What If? conlanging: Setting themselves very narrow questions, and then working through languages that answered that question. All conlanging has a bit of What If? to it, in my mind, but engelangers in particular sometimes raise this to be the most important aspect of their conlanging. Conlangs like this can often be summarized by a single sentence: What if English was reduced to only vocabulary derived from Greek? What if a conlang were used as an intermediary step in teaching monolingual English speakers foreign languages? What if there were a German-Spanish creole? All of those are actual examples from our seminar, which produced very interesting languages. Though I’m not much of an engelanger myself, I love seeing these sorts of languages: At their best, they’re perfectly focused, carefully planned, clever throughout. I tend to feel that most good artlangs start with questions like these, but then expand to fill more of the available creative space.
If it sounds like I’ve been a bit hard on my classmates, know that that is not my intention. I loved hearing about most of these languages every week; I loved watching them progress; I was genuinely inspired by many, and thought many others good enough to be worth continued effort. And seriously: This is exactly where first-time conlangers should be. I’m not trying to criticize, only to notice.
In sum: An excellent class. I hope to see many of these conlangs developed further. I would love to run a conlang feedback circle – any takers? 
We spent much of this class (as a wrap-up to our general unit on phonology) critiquing Klingon. Somehow, the general consensus in the rest of the class was that Klingon was a generally successful language, given its design goals. I’m not entirely willing to concede that. Keep in mind that I actually haven’t taken much of a look at anything beyond the phonology, yet – we’ll get to that in a later week (and I’m sure I’ll rant about it here, too). But I feel fairly strongly that the phonology of Klingon is a complete mess.
This was the list of presumed design goals we generated in class, more or less:
- Consistency (internal and with prexisting dialogue; i.e. don’t piss off fandom)
- Ease of learning (for actors, of course)
- Alien aesthetic
- “Harsh”, “angry” aesthetic
- Believable
That Klingon succeeds in not pissing off fandom is self-evident, of course. Ease of learning for actors is not quite knowable, from where I stand – without actually reading interviews with actors, etc., I have to assume that it succeeded well enough. But, with respect to the next three counts, I feel that Klingon fails somewhat miserably.
With regards to alienness: Seriously, uvulars and a retroflex consonants are the best you can come up with? Of course, a truly alien phonology wouldn’t rely on a human vocal tract, but even leaving that aside… Surely clicks, implosives, and (say) creaky voice are much more effective routes to “alienness”, for Americans. Just a few of these need to get added – too many will violate simplicity . But many Americans can produce at least the usual “tsk” click; creaky voice is easy to train.
Similarly with regards to the “harsh” phonology: Why avoid all consonant clusters? Surely, if we’re just going to go along with standard Western (/Tolkien) sound symbolism, a “harsh” language should have lots and lots of consonants (possibly even the syllabic ones that we don’t get in English). And do alveolar trills really sound that harsh? I’d think that Americans would mostly associate them with Spanish, which I’d think would have generally sanguine connotations…
But, of course, I think the biggest disaster is believability. When explaining why the /d/ phoneme should have a different place of articulation than the /t/ phoneme (it’s retroflex, actually), Okrand basically says, “Well, they’re alien, so they don’t have to follow the same rules!” Ok, sure, granted – but presumably they follow some rules! What motivation could there possibly be? Maybe voicing is partially linked to place of articulation throughout the language (some weird cognitive constraint, say). Well, ok – so why is it only /d/ that moves, not /b/? As another example, real languages generally try to find a balance between laziness and distinctiveness – and uvular sounds are enough harder to produce than velar ones that it would be extremely unlikely that the velar position would have no stops, but that the uvular would have one. Are Klingon’s vocal tracts just different enough to make velar sounds harder to produce than uvular ones? But then, why do we have velar fricatives, but no uvular fricatives (but a uvular africate)?
And I certainly can’t think of any justifiable reason to have the stress system of the language treat syllables ending in a glottal stop as extra-heavy…
So, no: I don’t find Klingon to be at all believable. Part of building a language is motivating the choices you make. In a language like Klingon, with a fictional species of speakers, this means justifying the choices with respect to the biology and culture of the that race; I don’t feel that Okrand has managed that.
And don’t even get me started on the ridiculous transliteration scheme…
The first week of LING 155 – a.k.a. Linguistic Typology and Constructed Languages, a.k.a. the Conlang Seminar, a.k.a. Fun 101 – is past and done. A brief report:
We spent the first part of class discussing chapters 1-7 of Okrent’s In the Land of Invented Languages, namely the discussion of philosophical languages. As a long-time conlanger, I’d encountered most of the languages discussed in this section before; it was moderately interesting to hear, however, the various reactions from the rest of the class. These ranged from wistfulness (“I know Wilkin’s Universal Language was impractical, but wouldn’t it be cool if we could make it work?”) to dismissal (“I can’t believe anyone actually tried this!”).
More interesting, however, was hearing the other (20) members of the class introduce their conlanging projects for the semester. Much of the class is effectively going to be a conlanging workshop, with students presenting their languages and critiquing each other’s. This week, we were just presenting initial (completely non-binding) sketches; these mostly ended up being about cultural/historical context, rather than language details (with a few notable exceptions). A few trends emerged:
Probably around a third of the class is working on a magical language of some sort. These range from Tolkien to Gaiman in magical inspiration, but share the notion that some aspect of the language is “true”, isomorphic to the universe, and capable of causing physical changes to the world when used correctly. At least one of these was specifically influenced by D’ni, as you might imagine. There was also significant overlap between the magical conlangers and the folks who expressed wistfulness about the great philosophical languages – the underlying desire being isomorphism between language and universe. My favorite of these sketches was a story about a culture trying to recover the ancient form of their language, as phonological change had obscured the language-universe isomorphism and rendered any spells not in common use ineffective. Just imagine: A fantasy adventure with historical linguists as the heroes!
On the far end of the spectrum from the magical-language people were the staunch non-conworlders. Probably another third of the class explicitly refused to have anything to do with historical-culture background for their language. In some cases, they set about creating grab-bag languages of their favorite features, sometimes with an overarching goal, sometimes without. Several, though, are working on either a posteriori conlangs or effective auxlangs. One student will be working on a version of English reduced entirely to vocabulary with Greek etymology – a fun twist on the typical “only Anglo-Saxon” trend with English. Possibly my favorite is an auxlang designed not for actual use, but rather as an intermediary step towards learning extant natlangs. Imagine encountering new grammatical possibilities – case, as an English speaker, for instance – in the context of a highly regular language, before you encountered all the complexities of actual language. Possibly my favorite idea for this language, which may or may not be outside the scope of what’s possible this semester, is to make it modular: You could teach learners a version of it extremely similar to their native language, and then start plugging in modules exhibiting properties not extant in their language, customizing it for whatever language(s) they eventually need to learn.
Proposals which did not fit into these two categories were the usual grab-bag of mixed art- and engelanging, my personal favorite genre (and, of course, the realm that my project for the semester falls into). Only one person is working on a non-human language – I was slightly unclear on the physiological details of the species in question, but they sounded a bit like sentient elephants – but was also going to be focusing on the various human dialects of this species’ language. One person is working on a language with the interesting morphosyntactic premise that all verbal afixes form a cluster somewhere independent of the verb (among other interesting ideas). One person is working on a bimodal spoken/signed language (inspired partly by Dritok).
I’ll be working on an as-yet-unnamed consign, which I’m very excited about. I’ve posted the initial design document I handed in this week in a separate post here.
All in all, this class is going to be an enormously fun way to spend my last semester at Swarthmore.