A critique of Loglan

By W. A. Verloren van Themaat

THE CONSTRUCTED LANGUAGE Loglan is intended to serve in two ways: for experimental testing of the Whorfian Hypothesis regarding the influence of language on thought; and as a ‘culture-free’ and maximally precise logical language that still is capable of expressing also nearly everything a natural language can—allowing exceptions in functional areas such as that of literature, requiring strong psychical identification of the user with the language. (Esperanto, of course, has produced a literature of considerable quality, but discussion here of potentiality for literature in Loglan is beyond scope of this critique.)

But in order to be significantly useful in testing the Whorfian Hypothesis, with regard to the influence of language on thought, a language must be spoken as well as written, however restrictedly and experimentally. Consequently, there are two principal criteria according to which Loglan can be judged: whether it is truly as logical and precise as it claims to be, and whether it can be adequately spoken as well as written.

J. C. Brown clearly perceives the essential difference between a natural language and a system of symbolic logic: in a system of symbolic logic new terms are introduced strictly according to definition; in natural languages they are also freely introduced by way of metaphor. For Loglan, Brown uses the second introductory route as essential for the functioning of a natural language. Also—in respects less essential for the functioning of a natural language—he does not always follow the conventional notational systems of symbolic logic: e.g. , the word order usual in systems of symbolic logic is VSO (Verb—Subject—Object; the verb is the predicate; the subject and the object are the first two arguments); the word order in Loglan is SVO, in agreement with six of the eight most widely-spoken natural languages, and in order to facilitate the formation of imperative and negative (Loglan 4, XVIII).

It appears possible to avoid completely constructional ambiguities, i.e. ambiguities due to uncertainty regarding the constituent structure of a sentence—e.g., “they are (flying planes)” vs. “they (are flying) planes”—or a sentence from which an expression is transformationally derived (“the love of the father”: does the father love or is he loved?). A strict proof of the absence of constructional ambiguities is not yet available, but very extensive computer testing of the Loglan grammar has not so far turned up a single example of constructional ambiguity. To accomplish this Brown has used elements similar in function to brackets in mathematical formulas, but here too he does not strictly follow mathematical notation systems. Also, constructional ambiguity is largely avoided because in predicate strings, predicates are left-associated, unless the opposite is explicitly indicated. Unlike in mathematical notational systems, brackets do not always appear in pairs.

The most interesting feature of the grammar of Loglan is that, in imitation of symbolic logic, it unites all content words: nouns, adjectives, verbs, largely also adverbs and prepositions—but not proper nouns—into one part of speech, the so-called predicates. The simplest form of a Loglan sentence consists of a predicate with its arguments, which correspond to subjects and objects in European languages. So, Brown demonstrates that those parts of speech differ only according to their functions in the sentence, not according to the kind of extra-linguistic reality to which they refer. Thus he undermines traditional Western metaphysics, which relied precisely on hypostasization of the parts of speech of the Indo-European languages (noun —> substance, adjective —> accident, etc.). The striking thing is that where, on one hand, he does away with distinctions among nouns, adjectives and verbs, on the other hand, he makes the distinction between common and proper nouns far more rigid than it is in Indo-European languages and than is justified by conceptual analysis. One reason for distinctions between common and proper nouns being less fixed in Indo-European languages is that so much depends on differing word conceptions: “God”, e.g., is a proper noun for a monotheist and a common noun for a polytheist.

Loglan shows in many respects a fruitful originality in relation to the usual systems of symbolic logic. Thus, in Loglan the simplest forms of the logical connectives—such as “and”, “or”, “if” and “not”—do not connect propositions, but predicates or arguments; sentences are connected by more complex connectives. This makes for greater concision without loss of clarity.

Brown himself admits (Loglan 1, pp. 232-233) that the logicality of Loglan can only be provisory, that Loglan—as any possible language with the same purposes—must be based on a certain conceptual analysis, which will necessarily be made obsolete by the progress of science. In this respect, Brown cannot be faulted. A curious example is given in Loglan 1 (p. 64) where the equivalence a —> b ≡ ˥aVb is ‘built-in’, in the morphological structure of the word for “if”. Now this equivalence holds in the most usual formal logic—classical logic—but not in intuitionistic logic, which dates from as early as 1908. One might object that Brown ought to have been better-versed in logical literature. Then he would have known of intuitionism and avoided this error. That this is not the heart of the matter must be recognized, considering that this mistake would have been unavoidable had Loglan been constructed prior to 1908.

A more serious flaw is Brownʼs treatment of tense. In Loglan the untensed predicate indicates a mere potentiality, whereas the tensed predicate indicates an actual occurence, e.g. Da cabro = “X burns” (is flammable), whereas Da na cabro = “X is burning” (Loglan 1, p. 47). His tense system is logically consistent—the objection that Loglan has no translation for “X will be flammable” (Loglan 1, pp. 79-80) doesnʼt hold water—but against his argument in Loglan 1 p. 52, it does not follow the order of observational simplicity. What can immediately be observed is that something is “burningˮ, not that it is “burnable”.

As to its grammar Loglan could certainly be used for communication among scientists, and perhaps even on a wider scale. However, prospects seem less favorable regarding composition of the lexicon. Though Loglan is not intended for general use as a Constructed International Language (CIL) but as a tool, rather, for experimental investigations, it is intended still to be spoken and written, and in such way has a certain similarity in purpose to CILs. Pursuing such a purpose, Brown should have consulted the history of the CILs, in order to know by experience what works in a CIL and what does not. Perhaps the most striking lacuna in the bibliography of Loglan I is the complete absence of books about the history of CILs. Brown shows a very defective knowledge of them. Such apparent ignorance of the work of predecessors, as so frequently happens, has led in this case to repetition of errors long known to experts in the field, and also to searching for answers to questions no longer troubling informed practitioners in the CIL movement.

In construction of the Loglan lexicon, Brown has striven for resolubility—i. e. automatic resolubility of spoken Loglan utterances into words—and cultural neutrality (Loglan 1, pp. 249 and 263), in order to make Loglan suitable for cross-cultural experiments, and to avoid as far as possible Loglan words bearing unintended connotations belonging to ethnic languages.

As to resolubility, experience has shown that constructed languages, at least with regard to present usage, are mostly employed in writing. This holds true even for Esperanto, the most widespread constructed language, and will certainly hold for Loglan in the first few decades of its existence. So the advantage of resolubility is very small.

In order to achieve resolubility, Brown has applied the following, i.a.: 1) An obligatory form of the primitive predicates cvccv or ccvcv (c = consonant, v = vowel); 2) The construction of complex predicates with mutation of the primitive predicates from which they are derived (e.g., morma = ‘kill’ from morto madzo = ‘dead-make’).

1) By this means Loglan returns to the rubric of the so-called mixed constructed languages whose best-known representative was Volapük, which turned out to be virtually unusable. These constructed languages in principle borrowed their words from the natural languages, but required a very specific phonetical consistency which often necessitated mutation of the words beyond recognizability. In some constructed languages, all words had to be monosyllabic, in others they were required to consist of series of alternating vowels and consonants, etc. Absolutely necessary for a constructed language for successful general usage is a supple adaptability for a great many purposes, which authors of such find it difficult to provide. The mixed constructed languages turned out not to have this adaptability, since new words from natural languages more often than not became unrecognizable through attempted adaption to the phonetical rules of the constructed language, and so could only be more or less arbitrarily introduced by some central authority. A striking example in Loglan is trime = “instrument” (Loglan 4, p. 282), for which the corresponding words in five European source languages closely agree (Chinese does not make an essential contribution); the Loglan word, however, is unrecognizable in its obligatory phonetic form.

The contemplated minimal usage of Loglan would certainly be by a small circle of scientists. And in science new technical terms are so often introduced that a language, in which new words could only be introduced by some central authority, would be entirely unsuitable.

It is true that Brown provides a method according to which, in theory, every user could make new predicates, but the resulting word would so often be unrecognizable that it could only be feasibly introduced along with a translation in one of the source languages; in such way Loglan would definitely cease to stand on its own feet.

2) Resolubility necessitates the mutation of primitive predicates in the construction of complex predicates, since if they were compounded without mutation, a compound predicate in speech would be undistinguishable from a series of separated predicate words (Loglan 1, p. 262). Brown should have been warned by the example of the natural languages. In some natural languages, minor adaptations occur in the construction of words from morphemes, e.g.: vowel harmonization in the Finn-Ugric and Turco-Tartar languages and some etymological compounds were contracted after their formation. However, elimination of entire syllables during productive word-formation does not occur in any natural language. This is easy to comprehend, since it would make it impossible to understand compounds by analyzing them into their constituents. The free combination of morphemes to words, which gives flavor to good Esperanto texts, is impossible in Loglan. Complex predicates could only be introduced together with their defining primitives. It is as if somebody telling jokes would have to interrupt himself continuously to explain them.

In order to make the lexicon culturally neutral, Brown has constructed most primitive predicates by combining letters from words in eight of the world’s most widely-used languages (by medialization: e.g., blanu = blue from English blue and Chinese lan).

But the experience of the Esperanto movement has shown that such neutralizing of the lexicon is to be avoided. Rather, Loglan words would do better to carry over connotations of words in ethnic languages, since the stylistic and semantic character of a language is not only determined by its lexicon, but also by its structure. As Brown rightly observes, Esperanto has the character of a European language. In its lexicon, Esperanto is chiefly Romance, but its structure is in many respects definitely non-Romance. The experience of the Esperanto movement has shown that Esperanto in the hands of non-Romance, and often even Romance authors, assumes a non-Romance semantical and stylistic character, with Esperanto words often taking on nuances of the nearest equivalents in the native language of the author, or a language well known to him, and not of their Romance source words. By analogy it may be expected that Loglan—which already has a non-European structure—even if it had a European lexicon, in the hands of e.g. a Chinese would assume a strongly non-European semantical and stylistic character, with word-meanings nearer to those of their Chinese near-equivalents than to those of their European source words.

In the best CILs, medialization has only been applied to already very similar words, e.g. Esperanto kuk-o from English cake and German Kuche. Extensive medializing in another among the better-known CILs, “Medial” of Weisbart (1923-1925), showed so many disadvantages that it was abandoned by its author. But Weisbart still restricted himself to European languages. The extreme medialization of Loglan, applied to a very heterogenous group of source languages, must often make words unrecognizable even for people with prior knowledge of all eight source languages.

Brown has been led by an extreme tendency to minimize the number of primitive predicates, in order to render all other notions by complex predicates. Esperanto, which in its original form showed the same tendency, was largely corrected by the spontaneous evolution of the language. E.g.: the word paf-il-eg-o ( = lit. “big tool for shooting”), introduced by Zamenhof, was replaced by the more international kanono. Now a word exactly after the model of paf-il-eg-o has been introduced in Loglan. Moreover, his minimalization of the primitive predicates has often led Brown to the adoption of dubious or illogical complex predicates, e.g., numsensi (= “number-science”) for mathematics.

Brown has not succeeded in finding a fixed method for the determination of the arguments of a given predicate. Therefore, this part of the lexicon reflects some more fortunate and some less fortunate decisions. A very fortunate incorporation of the findings of modern science was giving a number of predicates—such as the words for fall, lift, lower, over, weight, and heavy—an argument for the gravity field. Very dubious is the argument for the place of words such as hospital, restaurant and theater. The arguments of true and false for epistemology reflect a subjectivistic philosophy.

Conclusion: Loglan is a very interesting language, but mainly as an object of study for linguists, philosophers, and logicians. As a means of communication it is not far from complete unusability, and therefore the many significant results Brown expects from its experimental usage will turn out in practice to be illusory.

The irregularity of Interlingua

By W. A. Verloren van Themaat

FRANK ESTERHILL in his article “Interlinguistics: some further comparative aspects” (Eco-logos No. 80) incorrectly quotes me as having stated “Interlingua is more irregular than the Romance languages themselvesˮ. What I said in Eco-logos No. 78 was “In some respects it (Interlingua) is more irregular than the Romance languages themselves.” The particular example given in that instance of numerals, in my article “Relative merits of Esperanto and Interlingua”, does not stand alone:

1) The orthography of Interlingua (at least the one used in Interlingua-English Dictionary) is more irregular than that of Italian, since Interlingua has preserved ph, th, y, which have been replaced by f, t, i, in Italian.

2) The system of numerals in Interlingua is more irregular than that of French, since in Interlingua all ordinal numerals from 1 to 10 have to be memorized separately, whereas French derives all ordinal numerals except premier by the addition of ième to the cardinal numeral; in addition to second it has deuxième.

3) The derivation from verbal stems in Interlingua is largely more irregular than that of English, since Interlingua has preserved both the present and the supinum stem of the Latin verbs in its derivation, whereas English has preserved only the Latin supinum stem in many cases (Interlingua rediger—redaction against English to redact-redaction).

Of course, a constructed language which was, for the most part, more irregular than the Romance languages would probably be too useless to find adherents, but that it is more irregular than the Romance languages in some respects is bad enough.

W. A. Verloren van Themaat studied mathematics and philosophy leading to a doctorʼs degree in 1963 from the U. of Amsterdam. His article appearing herewith is based on J. C. Brownʼs Loglan 1: A Logical Language, 3rd edition, 1975, and Loglan 4 & 5: A Loglan-English/English-Loglan Dictionary, 2nd edition, 1975; both were edited at the Loglan lnstitute, P. O. Box 12458, University Station, Gainsville, Florida 32602.

SOURCE: Verloren van Themaat, W[illem] A. ‟A critique of Loglan,ˮ pp. 2, 3, 6, 8, 9; “The irregularity of Interlingua,” p. 9; Eco-logos [Incorporating Biophilist Magazine and International Language Reporter], vol. XXII, no. 82, 4^th quarter, 1976.

Two Poems
(‘The Volapükists’ & ‘The Esperantists’)
by Arthur Dawson Foote

“La Neĝa Reĝino” & “La Poeto”
de Herman Gorter,
enkondukis & el la nederlanda tradukis W. A. Verloren van Themaat

Review of Bruno Latour, Steve Woolgar, Laboratory Life
by Verloren van Themaat

“Leibniz, Couturat kaj la Teorio de Ido”
de Tazio Carlevaro

Letter on Sapir-Whorf discussions at LogFest ’89 and other topics

Annotated Bibliography on Language and Thought
by Ralph Dumain

International Language Review (issues listing + selected contents)

Philosophical and Universal Languages, 1600-1800, and Related Themes: Selected Bibliography

Esperanto & Interlinguistics Study Guide / Retgvidilo pri Esperanto & Interlingvistiko

Reflexivity & Situatedness Study Guide

Alireteje / Offsite:

Prologo al la nederlandigo de La Infana Raso
de Willem Verloren van Themaat, trad. Gerrit Berveling

Bibliografio de la nacilingvaj tradukoj de esperantaj literaturaĵoj
de Willem Verloren van Themaat

Whorfian linguistic relativism and constructed languages
by W. A. Verloren van Themaat
(International Language Reporter, 3rd quarter 1969)

Forward (March 1969)
by James Cooke Brown
(Acknowledges Verloren van Themaat)

Loglan 1: A Logical Language
Rev. 4th Edition
by James Cooke Brown

Loglan.org

Loglan - Wikipedia, the free encyclopedia

Lojban - Wikipedia, the free encyclopedia

Lojban site

Book Review of Loglan 1
by Bob LeChevalier and Athelstan

insult: La Lojban

The Emergence of Lojban Nationalism

A Working Bibliography (Sapir-Whorf)
by Robert Gorsch

Lojban in Perspective by Todd Moody

Willem Verloren van Themaat - Vikipedio

W. A. Verloren van Themaat @ Ĝirafo

‘Hymne aan het leven’: Herinneringen aan een collega /
Jan Gielkens
(en la nederlanda / in Dutch)

Lojban revisited by R. Dumain

Lojban and Esperanto: Extracts from from ju'i lobypli
(including this article from #13)

Some comments on the Sapir-Whorf hypothesis and Loglan,
with "Further Remarks on the Sapir-Whorf Hypothesis and Other Matters" &
"Bibliography with Commentary in Progress on Language and Thought: Part 1"
by R. Dumain
(Why Lojban: Extracts from ju'i lobypli #6 - 8 / 1988)

The Alembic, no. 2, Summer 1989
(Includes "Retorts" by Mark Tierisch on Lojban
& R. Dumain on childish anarchism)