Speedwords morphemes are not self-segregating

"Speedwords morphemes are not self-segregating," wrote Rick Harrison in his 'Language profile: Speedwords'. Bob Petry has pointed out to me on several occasions that Rick is in error in the examples he gives. Indeed, Bob is correct. But Rick is by no means wholly in error and the basic point he made is, in my opinion, valid. Certainly it expresses in essence what I actually experienced some 50 years ago in practice, as I explain below.

Are nob, por & kony really ambiguous?

First let me deal with Rick's error. It is, in fact, not true that for one who is familiar with Speedwords nob might be either the root word nob = 'noble' or no + b = 'noticeable'; similarly it is not true that por might be either the root word por = 'important' or po + r = 'successor'.

It is a rule of Speedwords that such derivatives must not be formed if the written result co-incides with a root word. Most affixes (though by no means all) are watered down forms of root words and in the cases where a derivative, such as the two quoted in Rick's profile, co-incide with a root word, the affix must take the full form.

In "Dutton Double-Speed Words Companion to Text-Book" (1943) Dutton (who called his language variously "Dutton Speedwords", "Dutton World Speedwords" and "Dutton Double-Speed Words") explicitly wrote:
"Care needs to be taken that [it] does not cause confusion with an existing radical. Thus the word 'maker' must be rendered maer so as not to conflict with mar - 'marriage', and a 'judge' must be rendered juer - 'law'. (Maer and juer are of course pronounced 'mah-yerr' and 'zhoo-yerr', with consonantal 'y')."

Where there is no root word allomorph of the affix, if the affix does have two bound allomorphs then use is made of this to preserve the rule that a derivative must not have the same form as a root word. Speedwords has an "opposite of" affix used rather like Esperanto's mal- except that in Speedwords it is a suffix, being -o after a root ending in a consonant or -x after a root ending in a vowel, e.g.

• nor "north" → noro "south"
• de "day" → dex "night"

• BUT
• a "to" → ao "away",    thus avoiding confusion with ax "ask".

Thus quite clearly 'successor' must be poer and por means only 'important'. Similarly, 'noticeable' must be noib and nob means only 'noble'.

Is Rick correct about kony which he claims may be either ko + ny "to come near, approach" or ko+n+y "to cause not to come"? Only as regards the former derivation; indeed, Dutton himself gives kony in his 1951 Dictionary as ko + ny "to approach". It cannot be the latter, as ko+n is forbidden by the fact kon is a root word meaning "accord, agree". In any case, "not to come" would normally be rendered n ko.

However, in theory we could have kon+y "to cause [someone] to agree". There is no rule which forbids a derivative being the same as a compound of two root words. Indeed, as I shall show below, Dutton himself actually lists such derivatives in his 1951 Speedwords Dictionary.

So Bob Petry is strictly correct in saying that Rick is in error with some of the words he quotes. But the morphemes can be segregated with surety if and only if:

1. you have complete familiarity with Speedword's 491 root words and its battery of 20 affixes, and
2. the word contains only one root morpheme ('radical'), and
3. the root morpheme does not have the past particple suffix. (See below)

Of these three conditions, #1 is essential. With some words, if the first condition is met, there may then be no ambiguity about the morpheme division; but with many others there will still be ambiguity as to the segregation of morphemes.

The Experience of a Learner

For someone learning the language, it is a very different matter. Rick is quite right in that a learner who knew no "notice, observe" might very well assume nob meant "noticeable" if she came across it for the first time in reading, and then wonder why things didn't seem to make much sense. This is precisely the sort of thing that I experienced when I was learning Speedwords in the 1950s.

For example, having learnt that am means "love", and finding ame (love-AUGMENT) "to adore", ami (love-ASSOCIATION) "lovely", amo (love-OPPOSITE) "hate", a learner may quite reasonably assume amu is another derivative, i.e. love-FAVORABLE, meaning something like "affection". Reasonable the learner may be - but she is again mistaken. It is a separate root word meaning "amuse".

Having learnt od "manner, method, mode, way", our learner, meeting odo would reasonably assume it was 'method-OPPOSITE' and wonder if it was something like "chaos". No - appearances are again deceptive: it is a root word meaning "odor, smell".

When I read in one of my lessons that "maker" should be rendered as maer and not the expected mar because the latter meant "marriage", my immediate response was: "Wouldn't it be better to use a different word for 'marriage'?" I feel that the 'exceptions to the rule' Dutton had to resort to in order to avoid derivatives being confused with root words is a clumsy expedient. In fact there are, in my opinion, too many instances where we have to use er rather than the suffix -r because of "confusion with an existing radical". For example:

• pu "think", but puer "thinker" to avoid confusion with pur "pure"
• ce "receive", but ceer "receiver [person]" to avoid confusion with cer "certain, positive"
• do "dwell, reside", but doer "dewller, resident" to avoid confusion with dor "sleep"
• no "notice, observe", but noer "observer" to avoid confusion with nor "north"

Inbuilt Ambiguity

Sometimes indeed it is impossible to avoid derivatives which co-incide with root words. For example, the perfect passive participle of ra "work" must be rad even though it is the same as rad "root"; similarly the perfect passive participle of pa "pay" must be pad even though it is the same as the root word pad "pad".

Also Dutton is quite happy to pile on suffixes, even if the result could be read as a compound of two root words, for example:

• ryat might be thought to be ry "building" + at "anticipate, expect" meaning, perhaps, "research laboratory" or "arrivals and depatures building at an airport" (Speedwords compounds are often somewhat idiomatic).
  It is nothing of the sort; it is 'building-UNFAVORABLE-DIMINUTIVE' = "shed, hut"
• ravu might be thought to be ra "labor, work" + vu "see, sight, vision" meaning, perhaps, "landscaping" or "optical task".
  Again it is nothing of the sort; it is 'work-ASSOCIATION-FAVORABLE' = "calling, vocation".

Conclusion

So in short, while Rick may have been technically incorrect with two or three of the examples he gave, he is still, both in my opinion and in my experience, quite correct in saying: "It is also worth noting that Speedword roots and affixes are not self-segregating.......and words such as itollis and evue which combine morphological ambiguity with semantically ambiguous affixes, are difficult to analyse."

Even if you do have complete familiarity with Speedword's 491 root words and its 20 affixes, you can not be certain how the morphemes are segregated in either itollis or evue. Nor, in fact, even when the constituent morphemes of each word is known, the meanings of itollis and evue are not immediately obvious.

In my opinion, Dutton's methods for dealing with root+affix formations that would otherwise co-incide with existing root words put an unnecessary extra burden on the learner. Fifty years ago they seemed to me to be a fudge that ought to have been avoided. Nothing since that time has convinced me otherwise. In any case, there are examples, as I have shown, where root+affix forms cannot avoid co-inciding with root words; and when we compound and pile suffix upon suffix, we enter a veritable minefield of morphological and semantic ambiguity.