De Wahl's rule

The rule was published by Edgar de Wahl in 1922. In its classic version, it is very restrictive. It does not allow the formation of words such as facte (= fact), factor, from far (= to do); scriptura, inscription from scrir (= to write), or actor, action from *ager (= to act). For this reason, numerous proposals have been made to generalize it. As early as 1922, Abbé Creux proposed a version that in many cases much better reflected the actual phonological processes. In his version, after subtracting -er or -r:

1. c, g, h were changed to ct,
2. ig, ic were replaced with ect,
3. b was changed to pt,
4. m was replaced with mpt,
5. t remained unchanged,
6. t was added after the remaining letters.

Creux's rule was in many respects a significant improvement over de Wah's rule. It allowed for much more natural word formation. For example:

• ager –> actor, action,
• scriber –> scriptura, inscription,
• facer –> facte, factor,
• consumer –> consumption,
• deducer –> deduction,
• diriger –> direction,
• eliger –> election.

Despite this, de Wahl did not approve of the rule. He criticized it as complicated and not very useful:

I think that the ingenious rule of the esteemed Mr. Creux is too difficult for an international language intended not only for educated people but also for those of lower education. And the rule given by Mr. C. is very complicated, and its result is not as valuable as Mr. C. thinks, because the infinitives of the current Occidental are only less Latin, but no less natural^[3].

The rule generally had no exceptions and therefore produced many unnatural forms. In his response, de Wahl drew attention to forms such as:

• immerger -> immerction instead of immersion,
• mulger -> emulction instead of emulsion,
• disponer -> dispontion instead of disposition,
• opiner -> opintion instead of opinion,
• comprimer -> comprimption instead of compression.

In 1958, several years after the publication of Interlingua, Cosmoglotta No. 203 published an article entitled "The Problem of de Wahl's Rule" (Li problema del regul de Wahl), presenting a modification of de Wahl's rule by an author using the pseudonym Interlinguisticus . According to Andreas Künzli, author of the book Universalaj lingvoj en Svislando, this was the pseudonym of Hans Homolka, an Austrian Occidentalist. In the article, Homolka proposed a set of rules for acquiring the second verb root that significantly approximated Occidental to Interlingua. The proposed rule was very complex. It consisted of seven parts (items a to g) describing individual types of alternations occurring in Interlingua, along with exceptions.

According to Homolka's rule:

a) If after removing -er or -r the root ends in a vowel, we add t.

b) If it ends with c or g, we change it to ct.

c) If the root ends with d or r, we change it to s.

d) The fourth rule described changes to the root in a few cases where it ended in m or n:

• primer -> press,
• sumer, imer, contemner -> sumpt, empt, contempt,
• tener, venir -> tent, vent,
• poner -> posit.

e) The fifth rule described cases in which the stem ends with v or rb:

• recever, scriver -> prescription, script,
• sorber -> sorbent,
• solver, volver -> -ut-
• mover -> mot.

f) Rule six described further exceptions:

• mitter -> miss, peller -> puls.
• flecter, necter -> flex, nex.
• verter, sentir -> vers, sens.
• merger, rumper, avrir -> mers, rupt, apert.

g) In other cases, removing the ending gives the desired root.

Points a) - c) had many exceptions such as:

• leer -> lect,
• struer -> struct,
• liger -> lect,
• rediger -> redact,
• cognoscer -> cognit,
• miscer -> mixt,
• attender -> attendant,
• intender -> intent.

The Cosmoglotta's article from issue 203 mentions a rule proposed by Reeve:

The latest manifestation of this new interest is the article "De Wahl's Reformed Rule", published by our colleague W. Reeve in International Language Review No. 3. It is more complete than Creux's proposal, and its explanation shows that it is the fruit of serious study. It is true that it does not exhaust all the possibilities of "naturalization", but its author deliberately renounced this, fully aware that absolutely complete regularization would overcomplicate the structure of the language for only small real benefits^[4].

According to the article, Reeve's rule was very similar to Homolka's rule, but it did not contain exceptions and restored roots from Latin that do not occur in modern Western European languages^[5].

Contemporary attempts are also being made to generalize de Wahl's rule. One of these is Stief's rule. According to it:

1. You need to remove -er or -r from the infinitive.
2. If the remaining core ends in...
   1. -s , -t , -x - nothing changes.
   2. -d - we replace -d with s.
   3. In other cases we add -t . -b /-g becomes devoiced to -p /-c.

There are 13 exceptions to the rule:

1. ceder → cess
2. currer → curs
3. graver → grav
4. herer → hes
5. morir → mort
6. mover → mot
7. nocer → night
8. opiner → opine
9. rebel → rebel
10. Seder → sess
11. sentir → sens
12. venir → vent
13. verter → vers

This rule is very similar to Creux's rule. It allows for much more natural word formation than de Wahl's rule. Examples:

• facer -> facte, factor,
• scriber -> scription, scriptura,
• ager -> actor, action,
• leger -> lection.
• proteger -> protection.

• Henry Jacob, The Preparatory Work for an International Technical Terminology reprint of The British Steelmaker, December 1945
• Henry Jacob, Occidental (1922) by Edgar de Wahl
• De Wahl's rule