There are two main channels of entry into language for an autistic, one through poor babbling and echolalia, the other mainly through writing. In both cases, social interaction is avoided. The first entry does not take into account meaning; the second does not imply vocal jouissance. This results in a surprising cleavage between two modes of expression.

Often, when they speak, a doctor confirms, autistic people do so with an atonic, mechanical voice, as if […] the musical part of the language was dissociated from meaning, as if they could choose between speaking without music or making sounds without meaning: raw meaning or raw sound, informative code or sensitive emotion, but never both articulated. (Hébert, 2006, p. 208)

The raw sound, resulting from babbling and echolalia, carrier of elusive non-communicable affections, unfolds in a verbose language, characteristic of solitary vocal satisfactions; while the acquisition of language through writing favors the appearance of a “raw sense”, of an “informative code”, of a factual language suitable for communication but separate from affections. This devitalized language is that which contributes to social adaptation. The latter is made up of signs, while the former is lalangue made entirely of S1.

Verbose language prevails in autists who present the most severe disorders; while the other main language, the factual language, presents a most elaborate evolution in Asperger autists. When Williams describes this separation, she specifies that the factual language, often called functional, is the result of the part of herself that folded to the education imposed on her, that is to say, it is taken by the discourse of the Other, while the verbose language belongs to her personal universe “completely cut off from the rest of the world” (Williams, 1992, p. 274).

What are the characteristics of the language “of accumulations of facts” (1996, p. 169) mentioned by Williams? Let’s start from two examples mentioned by Kantzas. In the first one, Jacques answers a question about the dream as follows:

I have slept has snored tonight woken up. Well, you laid on the bed covered the blanket when you have slept you have woken up you have got out of bed. He put on his pants, the shirt, the sock I put on the sandals I put on the underpants. I put the zip needle sewed stitching. (Kantzas, 1987, pp. 121-122)

Kantzas takes into account that the question on the dream, a word without concrete referent, for Jacques without meaning, requires an answer that calls facts, the events of the night, tangible and concrete elements. In the second example, Georges repeats a short story:

A fur trapper wakes up very early in the morning. He leaves home with his rifle and many dogs to go to the woods. When the dogs see the fox, they start to bark. The fur trapper then puts the rifle on his shoulder, shoots and kills the fox. Then, the fur trapper goes to the market to sell the fox fur, which is used to make garments with the fur. (Kantzas, 1987, pp. 121-122)

These successions of events, without observations, without affections, seem to contemplate a simple presentation of things, without the implication of the enunciative voice of the speaker.

Another observer of these phenomena emphasizes that such observations have essentially a “confirmation” and unintentional nature. Aubin, for example, could not say that the teacher had punished one of his classmates because he had been bad. This child’s observations, reports B. Donville (2006), “were limited to details”, mentioning the color of a classmate’s coat, observations that suddenly came from somewhere else, while he was not asked anything in particular” (p.68).  However, the subsistence of an element of soliloquy is taken into account in these observations.

He would just stand there in front of his mother to inform her, then he kept quiet without adding anything else. Nothing in his language related, he did not describe, he did not expect to deduce, in the best of cases specific considerations were obtained without him ever drawing any consequences. (…) (Donville, 2006, p. 68)

Those observations are very different from verbosity: they are inscribed in an effort to communicate, which is why they must be produced in the language of the Other. On the other hand, vocal jouissance is erased, while the verbosity is affirmed (Maleval, 2007).

One of the peculiarities of the factual language of autists, highlighted by all specialists, is the massive use of nouns, a linguistic category that simply expresses the existence of things. The anchorage in the concrete demonstrated, comes from the concern for using essentially just words which referent is an object identifiable in reality. Clara Park, an observer attentive to the language of her daughter, specifies:

Weather concrete or abstract, she immediately and absolutely understood all the words, those that expressed definable and comprehensible concepts in themselves: “box”, “cat”, “giraffe”, “rectangle”, “number”, “letter”. However, she was incapable of understanding words for relationships, which only find their full meaning in a situation where the human element plays a role. (…) (Park, 1972, p. 216)

In fact, there are many words that require the apprehension of a context and a relationship with other words in order for their meaning to be grasped. For instance, either smallness or greatness cannot be absolutely understood. In fact, since Saussure, we know that “in language there are only differences”: sign, signifier and meaning are only defined in systems of differential and interdependent oppositions. This suggests that what an element means only comes from its relationship with others, which implies a subjective work, an exercise of judgment to which Kanner’s autist does not risk. He is oriented towards a language that would describe the facts without having to interpret them himself. Therefore, his ideal would be a code that would constantly and rigidly connect words to objects or clearly determined situations; Nazeer explains:

It is not the complexity of a language what implies a problem for autists, in fact language is likely to help them, inasmuch as the more present it is, the less a word runs the risk of being polysemic. The more rules and structures there are, the less an autist should rely on his intuition and context. (2006, p. 26)

The ideal for them, he points out, would be “a meaning/a word”, that is, a language that would be reduced to a code, therefore, it would be completely constructed with signs.

Kanner’s autists, Grandin states, “are incapable of assimilating the slightest divergence regarding the images stored in their memory” (1997, p. 184). It happens that they create words that can only be understood if one has knowledge of the first situation that created them. During the development of the language of her son Thomas, Hilde De Clercq (2005) was surprised to find that he invented words to name objects that had the same function but did not exactly look the same. For example, he used different nouns to name a “bicycle.” “He had a “bicycle”, a “tractor”, a “wheels in the mud”, a “wheels in the grass” and a “small feet on the pedals.” Everyone found him very creative, but I had my doubts, says his mother, I thought he couldn’t do otherwise. When I said to him: “Go for a ride on your bicycle”, he did not understand because he only saw the “small feet on the pedals.” For him, the bicycle did not have a generalized meaning yet. It is the same logic that led him to name the glasses as follows: “the furthest”, “the milkshake”, “the glass of Boma” and “the dish of the day” (De Clercq, 2005, pp. 22-23).

Children, linguists state, quickly learn that one thing can have several names, that the dog is an animal, and they can be called Medor and doggy. It does not seem to be a problem for them. “They expect that a father will also be a Mister and he will also be called Pierre or Paul or ‘dear’” (Boysson-Bardies, 1996, p. 158). However, that is not assumed by an autistic subject. For them, whether the sign is sonorous or written, it remains a long time correlated to a certain experience. It is what explains that the formidable musical memory of such autist always runs into the same error inherent in the first hearing, or that a spelling mistake is always reproduced because it was present in the first text where he found the word. The difficulty of the autist to generalize lies in the rigidity of the sign stabilized by the persistence of the learning situation, so that it is not suited to be modified when the context changes. “If I learned something standing with a woman on a summer’s day, Williams says, the lesson did not evoke anything if I found myself in the same situation elsewhere with a man on a winter’s night” (Williams, 1996, p. 91). Therefore, even recognized in the language of the Other, factual language remains compatible with radical lack of understanding that result from taking information literally.

“One day I was lectured for having painted on the walls of Parliament during an excursion, says Williams, I promised not to do it again. Ten minutes later, I was caught painting once again this time on the school wall. I never had the pressure of having disobeyed or failed to keep my promise only for the pleasure of playing the clown: what was valid for the walls of Parliament, for me, it was not valid for the walls of the school. I hadn’t done exactly the same as I had before. That was it.” (Williams, 1999, p.115)

Factual language, which enables an autist to communicate, is also that which allows him to construct a shared reality. The primacy of the sign in this language leads to giving a privilege to isolated linguistic elements to the detriment of their contextual consideration; this is what studies of cognitive psychology find when they show that autistic children treat information abnormally, not paying enough attention to the structure of the whole and too much attention to small elements of this structure. That translates into what is called their hyper-selectivity, which often makes it difficult to distinguish the fundamental part from the accessory part. This suggest, as Kanner noted from this first article, that “the meaning of a word becomes inflexible and cannot be used with anything but only with the connection acquired initially” (Berquez, 1983, p. 256), for example, a bowl will only be a “bowl” in a particular place or situation with a precise color. Failing in context incites the autistic child to initially understand the meaning of the word, not placing it in fields of signifying oppositions, but connecting it in a rather rigid way to the designated object. The autistic subjects experience any modification in the thing/sign relation as a threat to their own safety.

When a specific referent does not exist, the autist often is forced to invent it. Thus, faced with too abstract concepts, Grandin strives to transform them into icons:

For peace, she says, I thought of about a dove, a calumet or peace pipe, or photographs of the signing of a peace agreement. For honesty, I was someone swearing, hand on the Bible, to tell the whole truth before a court. (Grandin, 1997, pp. 35- 36)

The connection of the sign to the referent makes it inappropriate to cipher affections, which are expressed differently in each one, which possess nuances, which are often fugitive and changing, and which are difficult to objectify. Autistic children, Williams specifies, are “secretly caught in a mutilated emotion”, “they have feelings and sensations, but these developed in isolation. They cannot verbalize them in the normal way” (Williams, 1999, p. 301). “The brain, Harrisson confirms, does not receive messages from the body, even though brain and body each work separately” (Harrison, 2010, p. 311). This results in a lack of spontaneity and personal initiatives that force autists to observe social behaviors in an attempt to solve the cut between their intellect and their being of jouissance. An Asperger autist uses a surprising formula to describe a constant work of reasoning cut from what she has lived: she permanently needs to orient herself in social life “to think her life and to live her thought” (Damaggio, 2011, p. 168).

Factual language of signs is separated from the color of the affections. All autists who undergo a considerable regression on their functioning note, like Tammet, that for a long time they did not understand their emotions. “They were things that happened to me, he says, that is all, they did not come from anywhere” (Tammet, 2006, p. 118). They emphasize their inability to grasp them and to express jouissance and emotional life through language. “I could say what I thought, Williams says […], but not what I felt […]. I was emotionally constipated.” To express herself, she had to take advantage of her ingenuity by chatting with a loose and fortuitous air, “any other precise way of doing, I would have stumbled at the obstacle of emotions” (Williams, 1999, pp. 88-89).

Initially, the emotions perceived by the autist are not interpreted by the sign. “My face got wet, Tammet says, following the death of his dog, and I knew I was crying” (Tammet, 2006, p. 165). It is later and in relation to the body’s externality that the emotion makes sense. It is exactly what Harrison describes: “Autistic people have emotions but they must import the meaning of their emotions from the outside in order to have conscious access to it” (Harrison, 2010, p. 241). This indicates that emotions are learned intellectually.

I want you to show me the emotions, D. Williams ask a family of friends for. […] Thanks to some lines and schemes, I live the angry scale, the happy scale and the sad scale. On these lines, they marked the low and high alternatives: tired, busy, irritated, agitated, upset, annoyed, and furious. They tried to show me how each state could be translated into a face or reflected in actions (Williams, 1996, pp. 161-162).

The comparison with the assimilation of a computer program is established insistently to highlight the acquisition of a ready-made knowledge that still remains separated from its feeling. “As computer files, Williams notes, you can mentally memorize, extract and interpret games of emotions” (Williams, 1996, p. 288-289). However, she notes the limits of this imitative process in terms of her subjective apprehension.

However, she highlights, the game is not linked to a real feeling and the emotion represented is not necessarily understood, except for the mechanism of the mode and sometimes the moment of its imitation. […] An idea is never a feeling; it is simply the memory or the memorized mental repertoire of its impression. (Williams, 1996, pp. 288-289).

Affections should be learned from external representations. But, Sinclair notes, “there is a difference between being aware of your own feelings and knowing how feelings are named. There is also a difference between having feelings and having automatic connections between feelings and expressions” (1992, p. 297).

The initial rejection of the autist, often mute in his earliest years, to the presence of the voice in the interlocution, leads him to a solitary appropriation of language, which occurs through the written language. Is it possible to incorporate a language separate from vocal jouissance? Certainly, Lacan answers. Ordinary experience, he points out, is that whatever the subject receives from the Other through language, it is received in vocal form. However, he adds:

the experience of cases that are not so rare, even when brilliant cases like Helen Keller’s are always evoked, it shows that there are other ways to receive language, besides the vocal way. Language is not vocalization. Look at the deaf. (Lacan, 2004, p. 317).

Would the autist learn language as the deaf? Mottron immediately distinguishes a difference: “the socialization of the deaf is done by gestures, he writes, the socialization of autistic people is done by writing” (2004, p. 149). The latter’s preference for a non-social approach to language is almost unquestionable. This is evidenced by the astonishing cases of hyperlexia of those who learn to read without having been to school and without their parents having noticed it.

However, their appropriation of language presents yet another major peculiarity. Temple Grandin described it very accurately:

I think in pictures. To me, words are like a second language. I translate all the words, spoken or written, into colored and sound films; they pass through my head like videocassettes. When someone speaks to me, their words are immediately transformed into pictures. (Gradin, 1997, p. 19)

She points out that another high-functioning autist functions in a similar way, keeping a visual image of everything he read or heard. This is Vieniamin Cherechevski, whose “prodigious memory” was studied by Luria for about thirty years.

When Vieniamin heard or read a word, it was immediately transformed into a visual representation of the corresponding object. This image was very vivid and remained solidly fixed in his memory. If Vieniamin’s attention deviated from this image, it disappeared; if he returned to the original situation, it reappeared. (Luria, 1965 [1970], p. 34).

If words constitute Grandin’s second language, she suggests that pictures would be the first language. Perhaps this intuition is not without foundation. The anxiety caused to autists by ceding the object voice (Maleval, 2009) tends, not only to make them initially mute, but also to incite them to cover their ears a little to the comments addressed to them, so images represent a much less disturbing channel for them to understand their world, because they are accessible to solitary appropriation.

Cognitive psychology studies show that autists “don’t learn language through communication”; that is why they argue in favor of:

exposure to non-communicative language vehicles, first of all to communication exercises focusing on typical communication prerequisites. This suggests exposure to the writing or repeated phrases in predictable contexts, since autists direct their own interests towards television, screens and printed material. (Mottron, 2016, p. 93).

Signifiers are taken immediately in a chain that organizes them; while the signs of factual language are initially incorporated one by one. Even though autists have reached certain organization of memorized signs, they also often describe their perception as “fragmented”. This characteristic is initially quite evident in children.

“In Amélie’s world it was clear that deduction did not exist, events did not happen, she never tried to ‘know why’ or ‘understand how’. Everything that happened in front of her had nothing to do with her, in fact she didn’t look for it, she didn’t even know it existed. When a story was read to her, she was only capable of accounting for the descriptive aspect of the image she saw, she remembered absolutely everything. The smallest details hindered her memory, but on the other hand she did not put in it any intentionality, no link of cause and effect between what the characters of the story did and the fact that such an event happened” (Donville, 2006, p. 76).

The Asperger autist no longer has a passive attitude towards memorized signs: he is now able to mobilize them to build local coherences. The appropriation of signs is initially made element by element. According to Harrisson (2010), memorization operates on the basis of “isolated data”, or according to Dawson, it is based on “a non-separable element of information” (Mottron, 2004, p.193). In individuals with the most severe level of autism, this element may be made up of words that are not linked, but for Asperger autists, it sometimes takes the form of a local coherence. The register of the elements is more complex than that of photographs: the perception of these themes, Mottron notes, “is not static”, it can be “multimodal (synesthesia) and allows manipulations on the memorized material” (Mottron, 2004, p. 186). It would seem, in fact, that the signs are arranged in sets organized according to each one’s own ways. The perceived coherence of the order of things and their regularities is one of them. Grandin uses another method: she relies on giving a main place to the chronological order [1],

to classify her “mental video library”; others resort to the image of a puzzle in permanent construction, others to that of a geographical map, or of a database, etc. It seems necessary to understand that each term occupies a place in a set, comparable to a file, which allows it to relate to other files, and that all this is sufficiently organized so that the subject can find the information.

Many autists describe remembering as reading a mental copy. Grandin proposes a similar analogy recalling “a computer program of graphic animation”, and specifying that its memory is constituted largely by a photographic storage of printed pages.

When I search in my head, I see the photocopy of the page. I can read it like a teleprompter […]. In order to find information in my memory, I have to play the video tape again. Sometimes it’s hard to find some data because It is necessary I try different cassettes until I find the right one, and that takes time. (Grandin, 1997, p. 33).

Of course, if the page has an error, the subject will repeat it systematically. According to the converging testimonies of autists capable of describing their method of functioning, they build “an internal database”, but the data are not “non-hierarchical”; that is to say, Mottron specifies,

that what is important and accessory is not implicitly distinguished. For example, the object and its context are not automatically separated and hierarchized in favor of the object. This is the reason why the change of context may cause a non-recognition of the object, or there may be a non-generalization of something acquired. (Mottron, 2004, p. 191)

Harrisson provides a very detailed description of the specificities of her memory. She models it as a “coherent fragmentation” with the following characteristics:

The register of each data is very precise and is done with a static structure. Each image is meticulously differentiated from the others. All details of each data are retained as they are perceived. Without any nuance. And each data is isolated at the beginning and then classified although it remains, at the same time, independent. The register is comparable to working on a puzzle. At the end, you see all the unique pieces that allow us to see the global picture. The pieces are associated but are not in interaction. The global picture of the puzzle (also known as the “geographical map”) is enlarged as the pieces are enlarged, depending on the number of pieces that can be linked, or the pieces to which a meaning is found to connect them with the others. These small global pictures formed by several fragments of information, very well defined in relation to each other, are merged into a large geographical map for each register. (Harrison, 2010, p. 91).

Michelle Dawson compares the arrival of information in the memory of signs “with the incorporation of a new piece in a puzzle under construction.” According to the intense memorization, and uselessness at rejection, she specifies that “she can’t leave it aside”.  Her puzzle is not the chaos to which are faced the most severe autists: she should incorporate every new element “in such a way that the puzzle remains as it was before this new piece, that is, it must be coherent in every possible aspect, direction and meaning” (Mottron, 2004, p. 194). Ouellette retains a similar picture, which he borrows from Atwwod, to describe the slow ordering of memorized signs operated by the high-functioning autistic person.

This work is compared to putting together a puzzle of a thousand pieces whose model is not on the box. With the passing of time some parts are assembled, but the global picture does not yet appear. Then, one ends up having enough “islands” of the puzzle to finally recognize the picture as a whole and all the parts finally find their place. (Ouellette, 2011, p. 181).

He is able to describe with great delicacy the main differences in functioning, distinguishing between that of the non-autistic subject, whose thought is ordered by the master signifier, and that of the autistic subject, who finds it difficult to articulate juxtaposed signs.

In a way, he points out, the neurotypical culture is undulatory, while the Asperger culture is corpuscular. Neurotypical thinking seeks to quickly, almost simultaneously, unite information from different senses and emotions to quickly construct a sense in the world for short-term benefit. […] The Asperger thought deals with this information rather slowly, hence the fact that individuals with Asperger are first interested in details rather than sets. Like a laser beam, their thinking focuses first of all on points. […] An Asperger individual acts like this to let the sense progressively emerge from the slow addition of details. (Ouellette, 2011, p. 184).

The above testimonies point to the fact that the classification inherent in memorized signs develops, not from a grammar or a signifying logic, but from images, or elements, and what unites them. “The continuation of my thoughts is not always logical, but often comes from visual associations” (Tammet, 2006 [2007], p. 85). Everything indicates that the same classification procedures govern the construction of a specific interest. Around the age of 10, on the occasion of the Olympic Games in Seoul, Tammet decided to become an expert in these games. He gave his teacher a file with hundreds of photos of athletes and events, as well as long lists of participants, results, statistics. However, he emphasizes that he organized the collage “according to an exclusively visual logic: athletes in red on the same page, athletes in yellow on another, athletes in white on a fourth, and so on” (Tammet, 2006 [2007], p. 85).  The advantage of this method is that it is based on an existing external classification that prevents the autists from facing the difficulty of creating it on their own. Not having the master signifier, it seems to be a visual classification that governs the organization of signs for the autist.

The discreet character of the sign only gives access to a fragmented memorization that some autists, thanks to an intense mnemonic work supported by the visual, are able to make more or less coherent. Non-autistic subjects do not have this difficulty. From the beginning, the elements that structure their thinking and their being are placed in structured sets. The signifiers, according to Saussure, are always trapped in a chain, they are connected in a synchrony governed by the laws of grammar, so that their learning spontaneously carries with it an assimilation of grammar. That is why linguists consider that non-autistic subjects are true “grammatical geniuses” between the ages of eighteen and twenty-four months. At that age the grammatical system that governs their language is fixed. They must carry out a complex ordering that implies a precise analysis of the phonetic segments of the words and their combination, as well as an order of grammatical data. They have to “integrate into the lexicon the phonological rules that control the pronunciation of words, and the morphological rules that govern their construction” (Boysson. Bardies, 1996, p. 218). However, it is remarkable that the grammatical genius comes to the subject by exposure to language and not by learning it. Lacan emphasized that this phenomenon makes an objection genetic psychology in its approach to the subject:

from his early manifestations of language, the child makes use of syntax and particles according to the nuances that the postulates of mental genesis should only allow him to reach at the peak of a metaphysical career. (Lacan & Cénac, 1950 [1966], p. 142)

Indeed, linguists agree that “the essential elements in grammar are known before they are taught, for they are an essential part of our capacity to express ourselves” (Boysson-Bardies, 1996, p. 218). Most of the studies show there are few correlations between the mother’s language and the child’s linguistic development. Parents “do not teach” the language to their children; they provide them with models (Boysson-Bardies, 1996). It is not learned like multiplication tables, and its acquisition does not require the intervention of the pedagogue. It brings preconscious mechanisms into play.

This grammar acquired without learning, inherent to the synchronous organization of the signifying chain, is clearly non-existent for the autistic child. The discourse confirming this uses very little of the terms like “because”, “since”, “therefore.” The acquisition of relative concepts, Peeters points out, is difficult for autists, “since words such as ‘big’, ‘small’, ‘broad’, ‘narrow’, ‘about’, ‘on the other side of’, ‘after’, ‘give’, ‘take’, take their meaning according to the context and with respect to the relationship they have with the other words in the sentence” (Peeters, 1996, p. 68).

In order to make an autistic child understand the concepts of “big” and “small”, he points out very precisely, “it should be possible to communicate its meaning to him starting from a “literal” perception: “this is small”, taken in an absolute sense, and here one sees the invariable sense of “big.” Unfortunately, this is impossible” (Peeters, 1996, p. 68).

What he calls “literal perception”, we have highlighted, is above all a visual and even tactile perception. Every autist perceives that words that cannot be related to an image are difficult to understand. On the other hand, it is well known that they have a tendency to take expressions literally, that is, not to contextualize them accordingly.

Due to the isolated appropriation of each sign, the autist looks for rules to organize the relations with each other. The main rule is found in the order of things, hence their invariable concern. Just as Kanner included the search for immutability:

Their world must seem to them to be made up of elements that, once they have been experienced in a certain setting or sequence, cannot be tolerated in another setting or sequence; nor can the setting or sequence be tolerated without all the original ingredients in the identical spatial or chronological order (Kanner, 1943, p. 263).

Breaking the immutability puts at risk the fragile external order of the signs. In this respect, Mottron posits a very close hypothesis. In the absence of a spontaneous link between the various memorized elements, he writes, the subjective universe of autists

is identified, until an advanced age, with a non-hierarchical set of perceptual, visual, auditory and linguistic memorizations. Therefore, its internal coherence depends on what is perceived. This would be one of the possible explanations for this universal taste of autistic people for structured environments: their sense of permanence would go through the continuous perception of regulation in the environment. (Mottron, 2004, p. 186) 

Behaviors of immutability represent rules that they have invented or regularities that they have observed. I functioned, Williams says, “under a large number of very strict principles” (1999, p.139). She did not ignore they were “her own rules” and that “they were not compatible with those of the carefree life of well adapted people” (Williams, 1999, p.139).  The following anecdote reveals that the rule is so valuable that respect for the rules can lead to absurdity and submission to mistreatment. When Gerland entered primary school, the boys told her:

“We are going to hit you once a day.” I thought it was a strange rule, she said, but I accepted. Because there were a lot of things happening in school that I didn’t understand and which I just had to accept. The boys asked me to follow them to the basement restroom, with direct access from the courtyard. There they punched me in the stomach every day, mostly just once. Maybe it wasn’t as much fun to hit me, because my pain threshold was very high and when it hurt, I never let it show […]. They hit me every day until someone told the teacher. I didn’t like that, she had nothing to do with me and I found it insulting. It was now quite clear that I have been deceived in some way, so I felt stupid. Hadn’t I gone and found those boys myself, in case on some days they had forgotten to hit me? I had thought it should be like this. (Gerland, 2004, p. 88)

The first sign acquisitions are made in a static way; however, when the capacities of the subject are developed, he gradually becomes capable of mobilizing them.

“The autist does not have access to the order on his own, unless he has reached quite a high level (once he has access to movement, interaction)” (Harrison, 2010, p. 85). It is then verified that he is capable of detaching them from their rigid connection to the referent. As the ordered signs interact they become less static and tend to become detached from the learning situation. They are no longer just terms-labels, she added, the subject organizes them his memory in his own way, allowing him to introduce them into relationships of mutual contraposition. When Grandin is forced to resort to icons such as “a dove or a pipe” to represent the abstract notion of peace, it is not only because of the link between these words and the image of a referent: she inserts them in opposing relationships for the dove with other birds, and for the pipe with other objects that are smoked. The initial fixation of the meaning of the sign tends to decrease progressively through intense memorization work. As a result, it provides a certain flexibility in its use. Not without resistance, the autist acquires the notion that the same object can be designated by different signs, that the location of the object is not rigid, even that it is necessary to distinguish the thing and its image.

In autists the mnemonic ordering of signs is organized according to a spatial logic: the models mentioned are the geographical map, the puzzle, a CD-rom storage and a compartmentalization fitted with labels. Structuring of memory is the result of a voluntary effort, it is organized from the visual, and is not driven by unconscious elements. The subject confirms to be in immediate presence of the stored information. The elements of his memory remain anchored in perception. On the other hand, memorization based on the signifier is not visualizable: it is governed by a preconscious grammar, as well as by the jouissance united in the unconscious to the master signifiers [2].

In the absence of spontaneous acquisition of the grammar inherent in the signifying system, autists are forced to learn it by heart. Sometimes they do it very well. Their mnemonic abilities are such that many Asperger autists have an excellent command of their language, or even several languages. They are not grammatical geniuses, so to compensate for the dispersal of signs, they tend to become mnemonic geniuses.

As the autistic subjects make a progress, their mnemonic faculties allow them to acquire increasing flexibility in the link of the sign with the image of the referent, also the sign capability to relate to other signs develops in such a way that the abstraction aptitudes of the autistic subject can be linked to those of the non-autistic subject. Therefore, there is question that cannot be left out: when the sign turns flexible and combinatorial, doesn’t it become a signifier? Should we agree with the Leforts that in autism there is “at least S2 because the Other speaks” (1998, p. 318)?  The signs that acquire a differential value are similar to the S2, but they cannot be completely identified: they differ in their difficulty in erasing their anchorage in a reference to the thing perceived.

The construction of reality and the social self is done essentially for the autist from signs that are difficult to detach from the ideal, that is, a word/meaning. The result is that better structured autists want to present a transparent, unambiguous social facade that conceals nothing from their thoughts.

There is no difference between my private person and my public person. In any circumstance, I am myself. I think that this authenticity, this naive transparency constitutes a quality.  However, it is part of my madness, perhaps because I expect the same transparency from others and, in that, I am wrong. (Ouellette, 2011, p. 19).

The autistic child struggles to do and to understand appearance: he finds it as difficult to lie as it is to grasp a false belief in another. The autistic child gives priority to statements that are based on the world of things and is closed to the enunciation dimension that implies a division of the subject.

A subjective mutation, which we cannot address here, occurs in high-functioning autists: the signs become less rigidly univocal, while a connection with affects is established. Verbose language and the language of signs tend to be substituted by a new language, that of the melting of affects.

[1] “my memories are always presented in chronological order and the images are always particular” (Grandin, 1997, p.29).

[2] “The unconscious has to do primarily with grammar, […] also, it has a little to do, a lot to do, everything to do with repetition, that is, the completely opposite side for what a dictionary is used for” (Lacan, 1971, p. 19).