Axon: Creative Explorations

 

Keywords: Ferdinand de Saussure - Glossolalia - Language Mechanism - Symbol - Physiology

 

Introduction

In May 1896 Ferdinand de Saussure became entangled in the first study to consider glossolalia a phenomenon of scientific interest. His colleague, psychologist Theodore Flournoy, was investigating the trances and glossolalia of a Genevan medium named Catherine-Élise Müller. Müller was, assuredly, a Francophone monoglot. Yet in her trance states she had additional ways of speaking that she called Martian and Sanskrit. In the case of the Martian, she claimed to be merely repeating the words of a Martian speaker, without understanding what she said. As for the Sanskrit, she believed that she had learned this language five hundred years earlier during another lifetime as a princess in India. Rather than dismissing Müller’s claims as absurd or insane, Flournoy saw these claims as hypotheses to be tested. If her Sanskrit speech contained any correct elements they should be accounted for in a way that was scientifically sound. Not being an expert in Sanskrit himself, Flournoy collected some utterances in transcript form and asked Saussure for assistance. It was not long before Saussure found that Müller’s glossolalia contained some Sanskrit and Sanskrit-like (his term: Sanskritoid) units that were assembled according to certain rules.

In this history there is an opportunity to learn more about psychological elements of language assemblage. Glossolalia remains a controversial area that few linguists will touch, probably because glossolalia speakers disown their speech, instead attributing their utterances to a deity, a spirit or an extra-terrestrial being. Even in the history of linguistics Saussure’s involvement in Flournoy’s study appears to have long been carefully avoided, though it did come under belated but heavy criticism from Tzvetan Todorov (1982 [1977]). Todorov argued that Saussure’s scientific resolve was weak, leading him to believe Müller’s spiritualist explanations, and consequently to miss an obvious ‘symbol’ at operation in her glossolalia. The symbol Saussure overlooked was a link between Müller’s avoidance of her mother tongue, French (français), and the absence of f in her Sanskritoid. This ‘symbol’ was so evidently the cause of the lack of f for Victor Henry (1901) who pointed it out only months after Flournoy published his study including Saussure’s almost verbatim analyses. Todorov writes unforgivingly, ‘[a]n (analytic) eavesdropper would have usefully replaced the practiced ear of the Sanskrit expert’ (Todorov 1982: 260). But in this episode we can find important beginnings of modern links between psychology and language. Specifically, Saussure’s analyses of glossolalia contain the roots of his concept of the language mechanism which today lies in the foundations of so many theories of human communication. The debate between Saussure and Henry is also instructive. On one side we have Saussure’s view that we rely upon unconscious structures to place sounds together in particular ways. On the other we have Henry’s view that speakers assemble speech according to symbolic stimuli. From another standpoint I will show that both linguists overlooked physiological elements of glossolalia that could much more simply explain Müller’s use of the sounds that made some of her Sanskrit assemblages almost convincing.

 

1. Saussure’s explanation of Müller’s glossolalia assemblage based on her Sanskritoid utterance, ‘atiêyâ…ganapatinâmâ’.

Saussure gave his most precise explanation of how he believed Müller assembled her glossolalia in the postscript to a letter to Flournoy dated June 19, 1896. In this document he discusses the language units of a particular Sanskritoid phrase and how they have been put together. The process that Saussure describes begins with an initial meaning that Müller wished to signify with her speech and plays out with a series of associative mental steps that occurred in an attempt to satisfy an unconsciously self-imposed ban on French, the only language she could speak fluently. The interesting thing to note here is that Saussure finds in a phrase that he ultimately refers to as ‘gibberish’, a mostly systematic assemblage of various fragments of language.

In the following passage ‘her’ and ‘she’ always refer to Müller, ‘the Sivroukian phrases’ can be read ‘Sanskritoid’ and ‘the Sivroukian state’ is merely a trance state, but specifically one in which Müller spoke ‘Sanskritoid’.

PS – Rightly or wrongly, I am now disposed to see in the Sivroukian phrases something analogous to the Martian interspersed only at intervals with Sanskrit shreds.

As a simple illustration of my idea, or rather my impression, suppose Simandini wants to say this sentence: ‘I bless you in the name of Ganapati’.

While she is in the Sivroukian state the only thing that comes to her is the idea not to say, or rather not to pronounce anything with French words. Nevertheless French words are the theme or substratum of what she will say, and the law that governs her mind is that familiar words are each rendered ​​by a substitute with an exotic appearance.

The important thing is: it must firstly and above all not appear to be French in her own eyes; and she is satisfied by saying, for example, Madame as Maguish or Mademoiselle as Manogadish, haphazardly inserting new forms in the place marked in her mind for each word in French.

It remains only to add (in my instinctive ‘explanation’) that sometimes the substitution is completely arbitrary and sometimes it will be influenced or determined by the memory of a foreign word, whether it be English, Hungarian, German, Sanskrit – with a natural preference for the idiom that best accords with the location of the scene.

That given, I try to tighten my gaze upon these hypothetical workings with an example of how a particular phrase is emitted: “I bless you in the name of Ganapati.”

1. Je. “Je” is forced to change. Does her memory provide her with an exotic word for je? No. Then one is chosen at random a = “je”. [Perhaps it has in fact been inspired by the English I, pronounced aï, but this is not necessary.]

2. Vous bénis [lit: you bless]; or bénis vous [lit: bless you] because if, for example, the word I was suggested by the English it may follow that the English construction occurs involuntarily in the words placed immediately afterwards.

Consequently we mark,

bless you

for: tyê yo.

The yo (you) may have been taken from the English you. The tyê = bless, could be taken from anywhere, as in Martian.

3. in the name of Ganapati: Of course the name of Ganapati is outside of any mechanism, and had to be taken from somewhere as it is. The remainder in the name of… that can be expressed with nâmâ, either by recollection of the German Nâme, or the revival of a Sanskrit nâma, also used sometimes. And finally the construction, opposing the order of French words will come on the wings of the German Nâme, after the German in Gottes Namen, in Ganapati's Namen.

This figuration is a simple process that I wish to show you. I attach no point of importance to the specific phrase I have used here.

In sum, the gibberish takes its components from where it can, and half the time is invented with the only rule not to let the audience suspect that it is French. Naturally, it is, I repeat, my impression, such as it is, that I give you. (O. Flournoy 1986: 193-194, my trans.)

The utterance ‘atiêyâ…ganapatinâmâ’ was only one of many ‘Sanskritoid’ utterances but this passage shows a process by which Saussure believed Müller unconsciously assembled them all.

 

2. Saussure’s explanation of the unconscious assemblage of Müller’s Sanskritoid.

Saussure incorporated Flournoy’s concepts teleological automatism and cryptomnesia in his discussion of the assemblage ‘atiêyâ…ganapatinâmâ’. ‘Teleological automatism’ is Flournoy’s term for purposeful unconscious phenomena. Müller’s Sanskritoid was her main evidence produced for the purpose of convincing séance attendees of her previous life in India (and therefore teleological), and this speech arose without conscious intervention (and therefore automatism). Cryptomnesia is Flournoy’s term for unconscious forgetting. He used it to describe memories that are incompletely recalled to consciousness; the subject lacks a clear reference point for how they have come upon this knowledge. When this occurs the subject experiences the recollection as knowing something for the first time. If, like Müller, the subject is spiritually inclined they might interpret this incomplete memory as knowledge from a past life or a message from a spirit-being. Both teleological automatism and cryptomnesia were fully integrated into Saussure’s analyses of Müller’s glossolalia. The parts of language he found in her Sanskritoid were all units that she had very likely encountered somewhere.

But Müller’s glossolalia posed the more intricate problem of how she assembled her linguistic knowledge into utterances. In her thirty years she had had ample opportunity to be exposed to a variety of languages. Her father was Hungarian and an accomplished polyglot, she had studied German at school, and she had an enduring interest in the orient and could have picked up several words of Sanskrit (such as names of deities like Ganapati) while reading books on that topic. To explain how these language elements came together, Saussure conceived of a mechanical process that could be performed without conscious thought and that followed a procedure something like a computational algorithm. First there is an underlying phrase constructed in French (e.g. ‘je vous bénis au nom de ganapati’). Next, French is set to ‘off’ preventing her from speaking the French words, yet her utterance is still motivated by the order of those words. A Sanskrit alternative is sought for ‘je’, and ideally she would simply replace each French word with a Sanskrit word.

This process of substitution runs into two complications, the first is a lack of Sanskrit vocabulary, and the second is the conflict between French grammar and the grammars of other languages. Müller did not know the Sanskrit word for ‘je’ but she did know it in English and manages to half pronounce the diphthong /aɪ/ before halting because English is not the preferred language (i.e. an obvious English word would be contrary to her claim that she was speaking Sanskrit). Nonetheless the half-pronounced ‘I’ introduces a second complication. The selected term brings with it grammatical rules that are contrary to the simple goal of substitution. With the utterance of ‘I’ the French SOV structure ‘je vous bénis’, has been exchanged for English SVO ‘I bless you’. Here the process is stretched to its limit and with nothing available to replace ‘bénis’ an exotic sound is uttered ‘tiê’ and followed by a return to English ‘you’, also cut short due to its inappropriateness. Next she could pronounce a Sanskrit word she did have in her memory, ‘ganapati’. But she had yet to find a substitute for ‘au nom de’. French does not allow this word order, but German (and her knowledge of German) does. The process ends with ‘nâmâ’. This was Saussure’s best guess at how ‘je vous bénis au nom de ganapati’ was spoken as ‘atiêyâ…ganapatinâmâ’.

In a discussion of Müller’s glossolalia, Boris Gasparov writes that Saussure ‘turned to analytical procedures that clearly overstepped the bounds of any orderly linguistic analysis’ (2012: 166).  Nonetheless, that disorderly linguistic analysis is of interest for the way it foreshadows Saussure’s later ideas about assemblage in language systems. Even in Müller’s apparently nonsensical assemblages, Saussure was able to find language systems at work, and in doing so he opened up her glossolalia to linguistic analysis. He saw that Müller was drawing on all of her knowledge of languages, making associations between them and expressing them in some kind of order. This analysis was Saussure’s first serious consideration of the psychological side of language. He grappled with the problems of association (e.g. the mental capacity to substitute one term for another) and syntagma (e.g. the impact of systems on the ordering of terms). Therefore the work that Saussure did in this unlikely context is indispensable in understanding his early thought on the central concept of general linguistics called the language mechanism.

There are two problems with Saussure’s analyses of Müller’s glossolalia. He points out - not in his discussion of ‘atiêyâ…ganapatinâmâ’, but later on - that Müller did two general things right. She did not use the consonant f at all, and she used the vowel a a lot. These are both features of Sanskrit phonology. Saussure observed these features of Müller’s Sanskritoid, and to be totally clear on the point that Müller’s glossolalia could be explained without recourse to spiritualist phenomena he would have done well to explain the absent f and the abundant a as elements of language Müller was likely to be familiar with. As for the a, he only suggested that Müller was imitating the few Sanskrit words she had encountered in Romanised script (all of which contained a high frequency of a). As this is an additional demand on the process he had proposed it is less satisfactory than simply drawing something from memory. But he offered no explanation at all as to why Müller was able to correctly exclude f from her Sanskritoid speech.

 

3. Observations of not-French and no f.

One of the rules in this process is particularly restrictive for Müller as a Francophone monoglot: no French. The importance Saussure attributed to this rule can be seen (in 1. above) in the way he specifies its purposes and refers to it again at the end of the discussion:

While she is in the Sivroukian state the only thing that comes to her is the idea not to say, or rather not to pronounce anything with French words. Nevertheless French words are the theme or substratum of what she will say, and the law that governs her mind is that familiar words are each rendered ​​by a substitute with an exotic appearance.

The important thing is: it must firstly and above all not appear to be French in her own eyes.

...

the gibberish takes its components from where it can, and half the time is invented with the only rule not to let the audience suspect that it is French.’ (O. Flournoy 1986: 193-194, my trans.)

Much later, Saussure found significance in the fact that, when she spoke her ‘Sanskritoid’, Müller never used the sound f. He wrote,

‘[F] is effectively a foreigner to Sanskrit

...

in the free invention of Sanskrit, we would have the chances of one hundred to one against in creating Sanskrit words using f, since this consonant appears as legitimate as any other if this fact is not known.’ (O. Flournoy 1986: 202-203 my trans.)

Saussure’s two observations, made three years apart from one another, remained separate in his analyses. On one hand, he had observed that one of the rules that ‘govern her mind’ is not to say anything in French. On the other hand, he had also observed that Müller correctly excluded the sound f from her Sanskrit-like speech. But there is a connection of sorts between these two observations.

Soon after the publication of Flournoy’s book Des Indes a la planète Mars (1900), which included the above observations by Saussure, Victor Henry (1901) saw how they might be linked. (Note that here Henry follows Flournoy in using the pseudonym ‘Mlle Smith’ to protect Müller’s identity):

If one general thought completely occupies Mlle Smith’s subconscious at the time she is assembling the sounds of Sanskritoid or Martian, it is surely that ‘French’ must be entirely avoided: her entire attention must be focused on that point. Now the word “French” (français) begins with an f and for this reason f must appear to her as the “French” letter par excellence, and thus she avoids it as much as she can. (Henry 1901: 23, my trans.)

Thus Henry gave a solution to a riddle that Saussure had left open:

Q: How did Müller know that f does not occur in Sanskrit speech?

A: She didn’t. Her exclusion of f was merely a symptom of her avoidance of français.

 

4. Todorov’s support for Henry’s connection, and subsequent dismissal of Saussure’s explanation of glossolalia assemblage.

Henry’s explanation acquires a lot of appeal and strength by showing that Müller could produce exactly the observed effect without any knowledge of Sanskrit at all. We can also commend Henry’s intuition for pointing out the connection between f and français very soon after Des Indes was published. Indeed, because it was so obvious to Henry, the fact that Saussure did not mention a connection between f and français has become a bewildering problem for some. This bewilderment is the point from which Todorov (1982) began his criticism. To understand his position we must keep in mind two assumptions.

1) Müller’s avoidance of French caused her exclusion of f.

2) The causal relationship between the avoidance of français and exclusion of f is obvious.

Henry’s remark suggesting a causal link between f and français seems to have immediately struck Todorov (1982) as both correct and obvious. He introduces a novel way of describing the operation that apparently took place in Müller’s unconscious mind with the term ‘symbolic’. The causal power of the symbolic relation comes out of its mental apparentness. Indeed, since f is obviously the first letter of français the proposition that there could be a causal relationship between them seems quite intuitive. After all, to say français one must begin by saying f. If this relationship is causal, it should be considered part of her speech-producing mechanism. In the term ‘symbolic’ Todorov suggests dream-logic and avoids baseless speculation that the rule in Müller’s mind of ‘no français’ entails a specification of ‘don’t even say the first letter of français’. In the world of dreams, Todorov seems to be suggesting, f is a sufficient representation of français.

But the purpose of Todorov’s argument is not at all to say that Henry successfully completed Saussure’s analysis of glossolalia by solving the one problem Saussure left open. Rather, starting from his platform of ‘the troubling absence of f’ (Todorov 1982: 258), he makes a sweeping attempt to dismiss Saussure’s analyses of Müller’s glossolalia by diagnosing Saussure with repression and accusing him of irrationality. This may seem like an ad hominem argument (at the time Saussure was irrational, therefore his analyses of glossolalia are rubbish) supported by a non sequitur (Saussure did not acknowledge the symbol, therefore he was suffering from repression), but it has been somewhat influential. Marina Yaguello (1991[1984]), Michel de Certeau (1996[1986]), and Françoise Gadet (1989[1986]) have repeated the same view with little or no variation. Saussure’s irrationality in this episode has also been glimpsed by Mireille Cifali (1994 [1982]), Tony Castle (1992), Daniel Rosenberg (2000) and John Grey (2011). Todorov lays out the main problem he sees in Saussure’s view as follows (again, ‘Mlle Smith’ should be read ‘Müller’):

[H]ow can we explain that Mlle Smith has guessed such a specific feature of the Sanskrit language without recourse to occult powers (since deception is ruled out from the start)? Would it have sufficed to glance through a treatise on Sanskrit to notice this? (Todorov 1982: 258)

Flournoy would have firmly answered ‘yes’ to Todorov’s last question. ‘Glancing through a treatise on Sanskrit’ was Flournoy’s explanation for all Müller’s Sanskrit knowledge (FIPM: 204 continued on 318-319 n. p. 204). This might be sufficient for Müller to pronounce Sanskrit words she had seen in a treatise. But to notice that the sound f was absent from all the words she observed and then to extrapolate from this observation a principle in the construction of the language system, would have required a level of astuteness difficult to fathom. Understandably, Todorov does not accept Flournoy’s explanation that Müller could have noticed such a specific feature in a glance. He finds that Henry’s observation solves the problem in a much more satisfactory way. In fact it is more than satisfactory for Todorov; he find’s Henry’s explanation to be correct. Saussure, despite being so close to making a connection, seems to have been unable to accept the kind of connection the data was suggesting.

Todorov’s application of the ‘symbolic’ introduces a way of thinking about causality in Müller’s utterances. He writes:

To make this discovery, one has only to recognize that the logic of symbolism is not necessarily the same as that of language; or even, more simply, that alongside language there exist other modes of symbolism that we must first learn to perceive. The letter f symbolizes “French” owing to a relation that is not characteristic of language conceived as a system of signs. (Todorov 1982: 259)

This is the most difficult part of Todorov’s discussion to follow because he claims that Saussure could have momentarily overlooked standard logic and his own theory of language as a system of signs to discover - with no objective evidence - that ‘f symbolizes “French”’ (ibid.). He next says that Saussure ‘does not admit diversity among symbolic systems’ (ibid.). Todorov means that there are other ways of thinking through this problem, but Saussure stuck so rigidly to his own narrow theory of signs that he was unable to see the spiritualist hazard in his path. In Todorov’s eyes Saussure would have done better to relax his theory of signs, accept that Müller’s exclusion of français causes the absence of f in her speech and work on the causal logic of that relationship later.

There is another line of criticism that Todorov reveals piece by piece. Todorov uses the limitation he perceives in Saussure’s analyses of glossolalia to insinuate that a restrictive theory of signs led him to obsess over unimportant details. If he had not worried about fine details he might have appeared uninterested in Müller’s claim that she spoke Sanskrit in a past life. But the details did concern him. Todorov says, ‘The analysis of the “Hindu” language seems to have fascinated Saussure to a degree that is difficult to fathom’ (1982: 256), and describes the great care with which Saussure carried out his analysis: ‘he attended séances and suggested possible interpretations of her case’ (ibid: 257). There may have been other ways to look at Müller’s ‘Sanskrit ability’, but Saussure sought more concrete explanations for particular details, and Todorov sees this level of work as unbefitting to the subject matter. Clearly a past-life explanation should have been out of the question, but Todorov senses that Saussure was taken in by Müller’s story-telling. Todorov sees the lack of the obvious symbolic explanation for the absence of f in her Sanskritoid as an indication that Saussure’s psychological state had been compromised, writing: ‘Repression where the symbol is concerned proves stronger than the prevailing scientific taboo ruling out recourse to the supernatural’ (Todorov 1982: 260). In other words, simply by looking at a problem in the wrong way, Saussure left open no other option than to concede that Müller’s glossolalia must be a spiritualist phenomenon.

 

5. Reasons why Todorov overestimates the value of Henry’s connection and why his dismissal of Saussure’s initial contribution is unwarranted.

Todorov grossly overestimates the value of Henry’s solution to the riddle left open by Saussure, as the only worthwhile thing to come out of the episode. On the basis that Saussure did not acknowledge this ‘symbol’ he accuses Saussure of obsession (1982: 256), belief in past-lives (260), impaired analytical faculties (260), outright failure (265) and a lack of imagination (270). Yet when Todorov upgrades Henry’s simple connection between f and français to the status of ‘symbol’ he offers no additional evidence that it functions ‘symbolically’. Another reason why Todorov overestimates Henry’s connection is that this ‘symbol’ only explains why a single sound is absent from Müller’s Sanskritoid. But Saussure’s theory has broad explanatory power of Müller’s glossolalia. Todorov’s position becomes even more tenuous when we realise that the historical impact of Saussure’s theory of glossolalia assemblage is a precursor to his later ‘language mechanism’ of the Cours. In other words, Saussure’s thought on glossolalia evolved into a concept of how all human speech is produced that remains critical to linguistics today (e.g. Moro 2008), but Henry’s addition to Saussure’s notes has, as Todorov admits, ‘no effect whatever on the evolution of science’ (Todorov 1982: 264). When we consider the explanatory and historical clout of Saussure’s theory of glossolalia Todorov’s position is truly puzzling.

If Todorov had shown that there may be other symbols at operation in Müller’s glossolalia, he might have established a theory to rival the explanation of glossolalia assemblage espoused by Saussure. For argument’s sake, let me propose another symbol.

Saussure also commented on the preponderance of a (O. Flournoy 1986: 204). If the absence of f was a symbol of Müller’s avoidance of français, then the preponderance of a might be a symbol of Müller’s desire to say a word beginning with a. The first vowel of the French word ‘Indes’ (IPA: ε̃d) sounds like the a Müller likely used in assemblages like ‘ganapatinâmâ’, ‘damasa’, ‘attamana’ etc. Since Müller’s Sanskrit-speaking past-life personality lived in India, the preponderance of a might be a symbol of her wish to say ‘Indes’ a lot.

Of course, not having been present at Müller’s séances I am not privy to the precise pronunciation of her a, and I do not insist upon this example. No doubt many such symbols could be discovered in Müller’s glossolalia, and if carefully explained they might constitute an alternative to Saussure’s theory of glossolalia assemblage, but not a replacement. Todorov is right to draw attention to the value of Henry’s original intuition as it importantly emphasises non-conscious operations in speech production. Henry had indeed pointed out certain significant aspects of Müller’s Martian that Saussure had failed to acknowledge. On this point Joseph writes, ‘Henry had outdone Saussure on the ingenuity stakes’ (2012: 456). But even Henry was aware of the limitations of his Martian ‘etymologies’, and Todorov goes too far when he elevates a tentative observation about Müller’s limited use of f in her glossolalia to a theory of the symbol.

 

6. A better solution to the riddle posed in Saussure’s observations and that Todorov believed Henry had solved.

There is a better way to explain the absence of f in Müller’s Sanskritoid which can also explain the preponderance of a. The first clue is that, like her Sanskritoid, Müller’s Martian is riddled with a and low on f. Henry notes only ‘six or seven occurrences of f in 300 words’ (Henry 1901: 22 my trans.). In a review of Henry’s text, Fred Corybeare suggested a partly physiological solution to the f problem, ‘I should rather ascribe it to a peculiar paralysis of certain muscles of the lips which may beset her in her abnormal states of consciousness’ (1903: 833). The simplicity of this explanation makes it more likely, but is there any need for this ‘peculiar paralysis’? In Henry and Saussure’s time there was very little data available to make useful comparisons between glossolalias, however, if such comparisons had been possible he might have noticed that the preponderance of a and exclusion of f was not particular to Müller’s ways of speaking. Goodman (1969) analyses glossolalia in four cultural settings. The data presented in this paper records a high frequency of a and not a single instance of f. Therefore, Müller has not been the only glossolalia practitioner to use a and f in this way. The reason for the exclusion of f is simply that it is relatively difficult to say, and the preponderance of a is explained by the fact that it is very easy to say. Symbols, and to some extent even systems are resolved in the principle of least effort.

 

Conclusion

Glossolalia is a controversial topic in academia, but the question of its assemblage is a real scientific question. Saussure’s commentary of Müller’s glossolalia offers a counter intuitive view of something that we think of as a poor imitation of language and shows that even this poor imitation contains structure and complexity. It is helpful because it offers an explanation of how she assembled all of her utterances. This episode in the history of linguistics also shows that speech assemblage is subject to so much we are not normally aware of; it brings together non-conscious factors of speech like association, competence and physiology. Müller often knew what she wanted to say, but because she could only formulate utterances with language units that were available from memory, Saussure was able to discover that her process of substitution relied on general phenomena. He analysed Müller’s utterances by identifying the systemic elements of the language units that went into their assemblage. It is true that Saussure did not find a symbolic operation in Müller’s utterances, and this has brought him criticism. I think this is an unfair criticism because there is no apparent reason for him to recognise the supposed symbol. There are other ways that the absence of f might be explained, and I have shown what I take to be the simplest. The study of glossolalia in linguistics remains controversial, yet in the years that Saussure engaged with the topic he was able to say about it something general and useful. More importantly, in his collaboration with Flournoy, a connection was forged between linguistics and psychology that would make way for the concept of the language mechanism and therefore the modern understanding of human communication. It may be that his lack of an explanation for Müller’s absent f means that his analyses of glossolalia assemblage are incomplete, but sometimes an f is only an f.