Jun. 8th, 2014

hrj: (doll)
Alpennian Language and Names: Part 1 - the linguistic background

I confess it: I have the Tolkien disease. I love using language as a worldbuilding tool. (I have entire trunk novels that were inspired by creating people to speak the langauges I’d invented.) Alpennia didn’t require that type of worldbuilding, but it was inevitable that I’d think a lot about the place of language, language usage, and more specifically about names in the world of my novels.

Researching language and names for an ordinary historical novel is fairly straightfoward: find the right sort of reference works and develop a philosophy about how you plan to balance authenticity and clarity. But while I placed Alpennia in the middle of a very specific time and place in our own world, I wanted to make it clear (in subtle ways) that it was something all its own and not a thinly disguised France or a Switzerland with the serial numbers filed off, or any other number of possibilities. I wanted it to have a look-and-feel all its own that still drew the proper connections with our own history.

I could neither let my imagination flow free and invent something from scratch, nor simply look things up in my over-stocked library of name resources and linguistic references. But both of those processes contributed to the end effect.

After drafting up this essay, I realized that it went on entirely too long, so I decided to break it up into three shorter bits: the background of the language, my strategy for naming characters, and the ways I use forms of address and reference to illustrate the structure of society and the relationships between the characters.

Alpennian Language

It went without saying that, given that I had invented a country, I would do something in the way of defining and describing what the language of that country would be. Geography alone determined the general parameters of the Alpennian language. Alpennia sits nestled in an imaginary space bordering on France, Switzerland, and Italy, roughly adjacent to the Savoy area (but without replacing any existing territory), and more pertinently, roughly in the region inhabited by the Langobard (Lombard) tribe in the post-Roman era. Based on this, we can assume that Alpennia was a Roman province at one point, that there was a strong Latin substrate and later invasion by Germanic tribes, and therefore that it ended up with a Romance language. That Romance language may have had some traces of an original Celtic substrate and definitely had a later overlay of a certain amount of Germanic vocabulary and personal names.

The primary reason to know something about the Alpennian language was to develop a distinctive approach to names, but there’s also a small sprinkling of vocabulary. Nothing where the meaning can’t be determined from context, but where I wanted the words to have an actual history and background. But, as I say, the most visible aspect of Alpennian language is the names. I wanted the reader to be able to trace connections to familiar names yet wanted to make them distinctive and clearly part of a unified culture. My method for coming up with “name seeds” will be described in the next section. I “grew” those seeds by applying a set of sound-change and spelling rules so that they would vary in systematic ways from the way names appear in neighboring cultures like French and Italian. And I wanted to come up with a target “look and feel” to guide my process and use as a touchstone for tweaking the results.

Based on the vague geographic target for Alpennia, I pulled out my copy of Bruckner’s Die Sprache der Langobarden which includes not only ordinary vocabulary but an index of all of the recorded Langobardic personal names. This was the touchstone for my look-and-feel. Now mind you, this is the look-and-feel of the names as recorded in the sub-Roman era. So we aren’t talking about a rational linguistic analysis, but of an esthetic target. I found that target in some of the spelling quirks of how those names had been recorded. In particular, I used the devoicing of stops in certain positions (i.e., b > p, d > t, g > k) as well as certain ways of representing sounds (“z” for the sound “ts”, “ch” for the sound “k” at the beginning of words). The idea wasn’t to imply that the spelling of Alpennian was static from the sub-Roman era through to the 19th century--certainly the specific names were put through major changes to produce the ones I used. But it gave me a target flavor that had some relevance to my fictional location.

After a little playing around, I came up with about a half-dozen sound change rules to use in generating modern Alpennian names. This is a far simpler system than would describe a real language over that time-span, but it was sufficient to the task. This next bit is going to use some technical linguistic terminology, but I’ll give a set of examples to show what I’m talking about.

Step 1: Clip off any inflectional endings, but keep a trace of any high vowels because they sometimes affect the vowel of the final syllable. (I told you I was going to use technical terminology.)

Examples:
Magdalena > Magdalen
Augustinus > Augustin
Bertolf > Bertolf
Johannes > Johann(es) (forms of the name occur based on both versions)
Stephanus > Stephan
Lombardi > Lombard(i)
Savatio > Savat(i)
castellanus > castellan

Step 2: Stress goes on the first syllable.

MAGdalen
AUgustin
BERtolf
JOhann(es)
STEphan
LOMbard(i)
SAvat(i)
CAStellan

Step 3: Certain initial consonant clusters are avoided, especially st-, sp- (as in French and Spanish), but also br- and similar. These are resolved by adding an epenthetic vowel (a schwa-like vowel which will normally be represented by “e”) either before s or between an initial stop and following consonant. This epenthetic vowel does not cause the stress to move. Consonant clusters involving “f” are also unstable, though not the rule isn’t as firm.

MAGdalen
Augustin
BERtolef (the final “lf” counts as a cluster because there is no vowel following, if it had been in the middle of a word the epenthetic vowel might not be necessary)
JOhann(es)
eSTEphan
LOMbard(i)
SAvat(i)
CAStellan (note that the “st” here doesn’t require insertion of a vowel because it isn’t a cluster -- it segments out as cas-tel-lan)

Step 4: All unstressed vowels are raised. (Initial epenthetic vowels generally remain as “e” but those elsewhere in a word may raise.) “A” may raise to either “e” or “o” depending on environment (I don’t have hard rules here so I wave my hands a little).

MAGdelin
BERtulif
JOhenn(is)
eSTEphen
LUMberd(i) (note that this is a rare example of a stressed vowel raising, no doubt there is a historic explanation that I haven’t yet invented)
SAvet(i)
CAStillen

Step 5: This step generally involves a variety of changes to consonants. Syllable-initial stops are devoiced as well as final stops. This does not apply generally to stops in clusters, especially when combined with liquids (l or r) but there are exceptions (I sort or wave my hands about dialectal differences at this point). As a rule, original “t” becomes “ts” (spelled “z”) except in consonant clusters (and certain other rare exceptions) however there are inconsistencies. V will sometimes devoice to F, but there are many exceptions. G when followed by another consonant may become a semi-vowel which generall shows up as i (turning the preceding vowel into a diphthong).

MAItelin
PERtulif (t remains due to the “rt” cluster)
JOhenn(is)
eSTEphen (t remains due to the “st” cluster)
LUMbert(i) (b remains due to the “mb” cluster)
SAvez(i) (note that v has not devoiced in this case)
CASzillen (this is a dialectal exception to the rule that t remains in clusters, I think there may be a sub-rule based on where the stress falls in relation to the cluster)

Step 6: This is the miscellaneous step with changes that aren’t quite as strict. Sometimes there’s syncope that collapses syllables, especially in the aftermath of epenthesis (resuting, in effect, in a vowel moving from one side of a consonant to the other). Clusters of very similar consonants may merge (e.g., “sz” > “z”). Original lost high vowels may result in diphthongs in final syllables, depending on my mood (e.g., Lombardi > Lumbeirt) or may be retained as syllables (generally as a schwa, spelled “e”). Esthetics are part of my process here. I'm looking for what changes result in something that “sounds right."

MAItelin
PERtulif
JOhenn(is)
eSTEphen
LUMbeirt
SAveze
CAZillen

Step 7: This step is pretty much just applying some spelling rules:
J >> I
PH >> F
TS >> Z (which I’ve already done in step 5)
initial K (as a sound -- it may be spelled C or K) >> CH (but K internally)
X (either as the gutteral “kh” sound or as “ks”) >> H

The initial CH spelling is to get a germanic feel without looking like German. The alternation is most apparent in the text in the shortening of Aukustin to Chustin, but this is the detail where my approach is most likely to mislead the reader and I will undoubtedly run into people who pronounce these names with an English-style “ch”. (And, of course, just to confuse the matter, Jeanne de Cherdillac’s title is pronounced in the French way with “sh”.)

Maitelin
Pertulif
Iohen(nis)
Estefen
Lumbeirt
Saveze
Chazillen

And that’s as much as you need to know about the language and phonology for the moment.

Profile

hrj: (Default)
hrj

July 2025

S M T W T F S
  12345
678910 1112
1314 1516171819
2021222324 2526
2728293031  

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Jul. 26th, 2025 08:46 am
Powered by Dreamwidth Studios