Romanizing Cthulhu, Part 2

Part 1

In a surprising development that no one could have possibly predicted, the inspirational well (and the fun) for this idle thought experiment ran dry very quickly. The truth can’t really be avoided; in their attempts to make “alien names”, Mythos authors mostly just replicated the sounds and forms of English words. Even those ludicrous consonant clusters just get broken up with epenthetic schwas if you try saying them out loud.

Ah well. I went into this knowing it was unlikely to amount to much.

Anyway, this post will contain all the meaningful thoughts I wrote prior to giving up.  With that done I can free up the mental hard drive space for the next fixation.

**

Now, before we start, I have a couple significant observations that could use expounding on.

Yi romanization isn’t viable on the whole 

As much as I would like to include it for novelty’s sake, I must concede that we’re using English orthography for the stop consonants. Saner that way.

The rarity of <e>
Outside of <ee> digraphs, the only places I have found it in the core HPL namelist are Gnopkeh, Y’ha-nthlei, R’lyeh, Yhe, Yeb, Nyarlathotep, Rhan-Tegoth, the nonce word “l’geb” from an incantation in Charles Dexter Ward, and a few locations in the Dreamlands.

Going by our earlier guidelines and dropping Nyarly, that gives us three /ɛ/, two /e/, and one /ɛj/

(R’lyeh is called “Relex” in The Mound, but I don’t have anything useful to pull from that at the moment)

That is an exceptionally small selection considering how many nonsense names we are dealing with, to the point where I am reasonably convinced that the mid front vowels don’t exist at all. I haven’t found evidence of a minimal pair yet to make the call.

(A minimal pair is when you have two phonemes in the same environment that, when swapped, would change the meaning of a word. A simple example would be pen vs pan (/pɛn/ vs /pæn/), which has the minimal pair of ɛ and æ.

I’ll be keeping my eyes peeled on this one, to see what I can cook up.

BACK TO THE LAB AGAIN

Picking up where we last left off, I’m going to focus on names that give us new rules or interesting new phonemes; the majority of Lovecraft’s names are actually quite tame and they don’t really have anything interesting to analyze

#3: Gyaa-yothn

A minor monster from The Mound, a story that will have outsize representation in this series (ed: It did not) for its higher-than-average percentage of weird names. This one can give us a couple of extra rules that are blessedly self-evident:

• HPL 7:  <aa> => /a:/
• Giving us a nice balanced triangle of long vowels.
• HPL 8:  CyV indicates that the preceding consonant is palatalized
• In this case, this is a straightforward  /ɟ/
• HPL 9: Syllabic consonants can serve as syllable nuclei
• This is going to be a huge help later on with those really absurd clusters. While we only have /n̩/ for now, /l̩/, /r̩/, and /m̩/ are likely.
• /-n̩/ is explicitly a pluralization affix in the story (no sign if it has any other function), and Yoth is the name of a place, so we have a surprising amount of grammar to potentially work with.
• Specifically, compound constructs seem to be built as [root]--[modifier] + [ending]; “gyaa-yoth” in the singular is something like “a gyaa, from Yoth” or “a Yothic gyaa”

#4: Yhe and Y’ha-nthlei

We get a double minimal pair here (at least according to our common sense rule of “if it’s spelled differently, it’s pronounced differently”): <y’h> and <yh> are not the same, and <e> and <a> are not the same.

<yh> is going to be a fricative due to HPL 6, and since the consonant component is the palatal approximate /j/ that’ll leave us with the voiced palatal fricative /ʝ/. Nice and easy. <y’h> can wait until we try to tackle apostrophes.

<ei> and <e> are going to be a pain in the ass, because in English these are treated as the same thing: we stick a /j/ glide after /e/ because English doesn’t like vowels sitting next to each other, but we stick that glide in even when there isn’t a vowel after it. But if we are going by what we’ve been given, /ɛ/ /e/ and /ɛj/ probably form a minimal trio in Aklo because they all show up in word-final position.

We’ll set that one aside for now. However the vowel is pronounced, <nthl> is a pretty obvious /n̩.θl/

#5 The god-damn apostrophes

Ah, the apostrophe. The bane of spec-fic since time immemorial, yet somehow they endure to the present day (See: local space-fascist frog-person Lē Zel from some video game)

In real languages, apostrophes can be used for:

• A glottal stop (ex. Hawai’i)
• A sign that letters or sounds have been elided (ex. cannot => can’t)
• An unwritten schwa (this is how the SFF apostrophe is typically pronounced)
• An ejective stop (this is how ejectives are written in IPA)
• An aspirated stop (as in Wade-Giles romanization of Mandarin)
• A marker to clarify if two sounds do or don’t form a cluster
• A marker to differentiate between phonemes that would otherwise use the same letters (ex. Breton uses <ch> for /tʃ/, and <c’h> for /x/ and /ɣ/
• A glottalized consonant

Since none of these authors gave a shit about any of this and just threw them into their alien names willy-nilly,  I’ll be using a similarly eclectic approach and give the apostrophe different realizations depending on its environment. Let’s look at our proper noun list and see what patterns we can find (limiting the survey to HPL alone for now).

• Y’ha-nthlei
• K’n-yan
• Thuum’ha
• T’la-yub
• Y’m-bhi
• L’thaa
• N’Kai
• K’thun
• S’ngac
• R’lyeh

Nearly all of these are word-initial C’C clusters preceding a vowel. There’s more variety in the strings of speech we get in The Call of Cthulhu  and The Case of Charles Dexter Ward

• Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn
• Y'ai 'ng'ngah, Yog-Sothoth h'ee - l'geb f'ai throdog
• Ogthrod ai'f geb'l-ee'h Yog-Sothoth 'ngah'ng ai'y zhro

The CDW lines have more to them, stick a pin in that for now.

Our simplest apostrophe is found in “Thuum’ha”, where it’s conveniently placed between two consonants at an apparent syllable boundary. That’s a pretty clear indicator that it is meant to designate that the syllables are split <thuum> and <ha>, rather than <thuu> and <mha>. It’d be helpful if we can find an <mh> cluster to compare it against, but even if we can’t I feel this is an easy one to pencil in under rule HPL 5.

Type 0 Apostrophe: used to denote either a glottal stop (when between two vowels) or a short pause (when between two syllables at a syllable boundary.

We haven’t seen the intervocalic version yet, but it’s such a common usage of apostrophes that I feel confident in including it.

Next up, we have our most common apostrophe usage, a word-initial C’C cluster. This is a really uncommon cluster in real-world languages (which is probably why it is so common in sci-fi), and our only real clue is that <y’h> and <yh> are differentiated.

Seeing no better option, I pencilled these in as click consonants, for the following reasons:

• Weirdo racists like those employed at Miskatonic would definitely consider clicks to be unknowable and alien, despite the fact that human toddlers can make them perfectly fine when raised in a language that has them.
• Weirdo racists like those employed at Miskatonic would probably brew up some bonkers ways of representing them.
• Most of these consonant-apostrophe-consonant clusters are word initial, and real-life click languages nearly always limit clicks to initial positions (usually the word or root, sometimes the syllable)

I have in my notes a sketch of how to determine what cluster equals what click, but like I said up top, the well is dry and there ain’t much to go on.

Type 1 Apostrophe: Word-initial consonant-apostrophe-consonant clusters are trigraphs representing click consonants.

Now, the bit for CDW that I stuck a pin in has one thing that could be interesting - the incantation is described as being “syllabically reversed” in its second half. That is, the syllables, not the individual sounds, are placed in reversed order. We end up with a pair of 'ng'ngah and  'ngah'ng; splitting that up we end up with ‘ng and ‘ngah - normally this would be a sure-fire sign of something like preglottalization, but I am proven a fool immediately by the next-door pair of y'ai and ai’y, where the apostrophe is clearly not tied to the syllable itself.

It was at this point that I got bored admitted defeat.

SO WHAT HAVE WE LEARNED

1. Sometimes I have stupid ideas that I shouldn’t entertain, but end up putting too much thought into anyway.
2. Science-fiction authors should have their apostrophe keys removed until they can prove they can use them responsibly.
3. Reality will always be stranger than fiction in ways that most authors will never be able to match, especially with language: Nuxalk contains many, many words that have no vowels at all, the textbook example being the grammatically correct but unlikely to be regularly used clhp’xwlhtlhplhhskwts’ (IPA: xɬpʼχʷɬtʰɬpʰɬːskʷʰt͡sʼ), which translates to "he had had (in his possession) a bunchberry plant".

HPL would probably spontaneously combust (in a highly ironic fashion) if he ever heard any of the Caucasian languages (that is, languages from the Caucuses). Dude had a mental breakdown over the existence of the Welsh; Ubykh has (well, had. Last speaker died in 1992) 84 consonants and two vowels.

Cthulhu ain't shit, phonologically speaking.