(Cross-posted at FacilitatedCommunication.org).
Do minimally-speaking autistic individuals have literacy skills?
You might assess this by seeing if such individuals can associate word labels with objects—placing the “dog” label, say, on a picture of a dog, and the “cup” label on the picture of a cup. You might assess this by seeing if the person can independently write or type out word labels of their own for the dog and the cup. You might assess this by presenting the person with a note that says, say, “Do you want to watch a video or go outside?” and seeing if the person, for example, responds by fetching a video-playing iPad. That note could be written on paper, but it could be typed out on the overwhelming majority of AAC devices that, contrary to what the pro-FC sector tells you, include keyboards and regularly expose users to printed word labels.
But why use such direct, straightforward literacy probes when you can instead deploy novel equipment and fancy statistics to generate much more indirect and highly indeterminate results? After all, that’s precisely what Vikram Jaswal did in his recent attempt (Jaswal et al., 2020) to find support for the variant of FC known as Spelling to Communicate (S2C, see Beals, 2021 for a critique of this highly flawed study).
This time around, in a study that has just come out (Jaswal, Lampi, and Stockwell, 2024), Jaswal sets out to explore the literacy skills of minimally-speaking S2C users—which, if they are found to exist in sufficient quantities, might convince some more people that S2C is valid. (In fact, the only procedures that would definitively test the validity of S2C, should anyone ever agree to participate in them, are a series of controlled message-passing tests).
In what the authors tout multiple times as a pre-registered study, they created an iPad game with arrays of symbols that flash one at a time in different sequences. Participants were measured on how quickly they responded to the flashing symbols by tapping on them. Some symbol arrays were alphabetic symbols (letters) that just happened to precisely replicate the letter arrangements of the letterboards used in S2C. Others were arrangements of symbols that look nothing like letters:
(From Jaswal et al., 2024)
In one condition, the Sentences Condition, the letters on the letter array flashed in sequences that spelled short sentences. In this condition the experimenter gave the participant an oral prompt about the given sentence, saying it twice, as in this example: “This time, the letters will flash in a sequence that spells, ‘I should water the backyard today. I should water the backyard today.’”
In the other conditions, no sentence prompt was provided. In the “Matched Symbol Sequences” condition, the symbols flashed in a position sequence that matched the position sequence of the flashing letters in the Sentences Condition. In the “Reverse Letter Sequences” and “Reverse Symbol Sequences,” the flashing letters and flashing symbols followed the flashing letters (or the positions of the flashing letters) of the Sentence Condition sentences in reverse order. For instance, in the case of “I should water the backyard today,” the “Reverse Letter Sequences” followed the sequence “yadotdraykcabehtretawdluohsi.”
Measuring the “latency,” or the time it took a participant to go from one flashing letter to the next, Jaswal et al. report the following findings:
Latency is, on average, lower (by several tens of milliseconds) for the Sentences Condition than for the other conditions.
In the Sentences Condition, but not in the Reverse Letter Sequences condition, participants tapped flashing two-letter sequences that are relatively common in English spelling (e.g., “he”) faster than they tapped flashing two-letter sequences that are relatively less common (e.g., “hs”).
In the Sentences Condition, participants showed sensitivity to words and word boundaries: they were slower to tap on the first flashing letter of the word than on the subsequent flashing letters of the word.
All of these behaviors, Jaswal et al. argue, mimic the behavior of literate individuals when they type. Just how closely, however, remains unexplored, as the study reported no comparisons on the flashing letter/symbol tasks between these participants and a comparison group whose literacy was beyond dispute.
But even if these patterns of relative tapping speed on flashing letters/symbols mimic those of people whose literacy is beyond dispute, is this enough to conclude that the autistic participants are literate? Does it mean, say, that they understand written commands to “water the backyard”?
In fact, the difference between
(a)
tapping the sequence “Ishouldwaterthebackyardtoday,” guided entirely by flashing letters
tapping that flashing sequence faster than both the flashing sequence “yadotdraykcabehtretawdluohsi” and flashing sequences of symbols that look nothing like letters, and
pausing slightly, while tapping the flashing sequence “Ishouldwaterthebackyardtoday,” between the flashing “I” and the flashing “s,” the flashing “d” and the flashing “w,” the flashing “r,” and the flashing “t,” the flashing “e” and the flashing “b,” and the flashing “d” and the flashing “y”
and
(b)
knowing what the printed words “I,” “should,” “water,” “the,” “backyard,” and “today” actually mean
Is quite vast.
Jaswal et al. acknowledge the multiple years of experience that most of their participants have had, via S2C, in pointing to sequences of word-spelling letters on held-up letterboards (two participants had between one and two years of experience; the remaining 29 had at least two). And they acknowledge that these experiences could have taught the participants statistical regularities about letter sequences and word boundaries that could explain these results. But they argue that this explanation is ruled out by the fact that the participants (1) weren’t faster with common, flashing two-letter sequences in Reverse Letter Sequences condition (where the letters didn’t spell words), and (2) weren’t consistently faster at tapping flashing letter sequences for more common words than for less common words.
There are major problems with these lines of reasoning, however. In the Reverse Letter Sequences, the sequences so obviously violated English spelling conventions that anyone who has grown accustomed to those spelling conventions—for example by being subjected to years of S2C—will quickly recognize that all bets are off. All of the sequences, that is, contained sub-sequences that are impossible to parse into words:
ohy, ehtr, dluohsi, afkaer, ofmaer, ahsdikeht, ohya, ehtno, tsti, sg, odtso, nahtre, ruo, tuo
If they had wanted to, Jaswal et al. could have easily avoided this confound by scrambling the letters in ways that didn’t result in such unparsable sequence. For example, instead of
yadotdraykcabehtretawdluohsi
they could have had
kaydotdrackabethawdlerthisou
Second, Jaswal et al. assume that what determines how quickly a flashing letter sequence is tapped out is how common the corresponding word is. But, as they themselves propose elsewhere, the issue is predictability: how predictable is the next letter based on what came before? Plenty of common words offer little predictability. Consider the words “it,” “had,” and “way,” which the authors found had high average latency-per-letter (slower typing). With “it,” common though it is, many letters could follow the “i,” and after the “t” the word is over. The same goes for “had” and “way.” It’s longer words with more predictable patterns, even if they’re less common, that should permit lower average latency-per-latter (faster typing). Consider the three words in Jaswal et al.’s sample that show the most extreme combination of lower-than-average commonality (relative to the other words) and lower-than-average latency-per-letter: “walking,” “raining,” and “outside”. By the time you get as far as “walki,” “raini,” and “outs,” the final letters are fully predictable (and, indeed, readily outputted by autofill functions).
(From Jaswal et al. 2024).
But Jaswal et al. propose that the patterns of latency observed in the Sentence Condition and not in the other conditions are evidence, not of a rote, statistical learning of letter patterns, acquired during years of S2C-based prompting, but instead of “motivated, literate participant[s]” who are able to “convert the sentence the experimenter spoke aloud into its written form.”
Of course, if Jaswal et al. had wanted to actually test this hypothesis, they could have included a comparison condition in which the experimenter refrained from reading the sentence out loud.
But this might have put a dent in their grandiose conclusions:
About half of the participants showed the signature pattern consistent with literacy; those who did not show the pattern may also know spelling conventions but be unable to demonstrate their knowledge on our task, which required sustained attention and speeded responses to over 550 targets. Regardless, our findings suggest that at least five times more participants have foundational literacy skills than would be expected from professional estimates based on nonspeaking people’s overt behavior. [Italics mine]
And:
Our finding that participants without phrase speech as well as participants with phrase speech showed the signature pattern of faster responses in the Sentences condition compared to the Matched Symbol Sequences condition… demonstrates that it is possible for nonspeaking autistic people to acquire aspects of literacy regardless of their speech proficiency. Second and relatedly, it is frequently suggested that more positive outcomes are associated with autistic children who achieve functional speech (i.e. phrase speech) by 5 years of age... At least as far as learning foundational literacy skills, there was no difference in our sample of nonspeaking autistic adolescents and adults between those who had some phrase speech and those who did not.
Turning to the question of how these minimally-speaking and non-speaking individuals might have acquired these literacy skills, Jaswal et al. end up, in an odd reversal, admitting that “statistical regularities clearly played a role” and that
the year or more of practice participants had communicating on a letterboard held by a trained assistant may have highlighted orthographic regularities and sound–letter correspondences, though we note that the training our participants had received on the letterboard does not involve explicit instruction on these correspondences.
Indeed.
But the authors appear oblivious to the implications. Learning “orthographic regularities and sound–letter correspondences” isn’t the same as being literate. Consider the stereotypical American Bar/Bat Mitzvah in which a teenager recites passages written in Hebrew without understanding most of the words he/she is pronouncing. Few people would say that that individual is literate in Hebrew. To be literate, sounding out and spelling out letter patterns aren’t enough: you must also know the meanings of whatever it is you’re pronouncing or spelling. Yes, orthographic regularities and letter-sound correspondences are included in foundational literacy skills. But if that’s all you have, you aren’t literate and, unless circumstances change dramatically, you never will be. This is especially the case if most of your “communications” involve S2C, which, with its exclusive focus on letter positions on the S2C board and on prompting kids to point to them (“right next door,” “keep going, keep going,” “you’re almost there”), completely omits meaning.
But Jaswal is no literacy expert. As a result, he has no problem closing the article with what reads like a subtle infomercial for S2C, complete with citations from pro-FC researchers:
An important question for future work will be investigating why, given this underlying orthographic competence, most nonspeaking autistic people do not learn to express themselves in writing. As suggested earlier, part of the challenge is lack of opportunity because of underestimation of nonspeaking autistic people’s capacity to acquire literacy…
[Note the omission here of the various ways, mentioned in the opening paragraph of this blog post and seen accompanying image, in which the evidence-based AAC devices routinely used with nonspeaking autistic people are in fact designed to promote literacy.]
The significant proprioceptive and motor challenges that many nonspeaking autistic people face can also interfere with learning to write by hand or type using conventional interfaces (e.g. Gernsbacher, 2004; Leary & Donnellan, 1995; Torres et al., 2013).
[Note the FC-friendly take on autism as a motor disorder that disrupts “conventional”—i.e., independent—typing, asserted in the pro-FC literature but never actually demonstrated.]
Thus, another important area for future work will be leveraging advances in technology to create accessible and customizable communication interfaces as well as opportunities to practice the skills required to use them.
Stay tuned for future articles by Jaswal et al. on how “advances in technology” can promote sophisticated literacy in minimally-speaking individuals, with the usual conflation of evidence-based AAC technologies with communication tools that, however high-tech they may be, are ultimately facilitator-directed and facilitator-controlled.
Meanwhile, a segment within Jaswal et al.’s final paragraph reminds us of the giant elephant that has been lurking in the room the entire time:
According to the nonspeaking autistic participants in this study and their families, the most effective means of communication available to them currently involves spelling words and sentences on a letterboard held vertically by an assistant. (All the study participants are working on the skills to be able to type without assistance.) Were it not for concerns about the role of the assistant…this would serve as prima facie evidence that they are literate. We used an experimental task that did not involve an assistant (i.e. no one held the iPad or touched participants) to investigate if these participants possess foundational literacy skills. [Italics mine.]
If the participants in the study were able to type out sentences without an assistant on stationary iPads with flashing letters, then why, when typing sentences that aren’t guided by flashing letters, do they need the letterboard to be held up by an assistant?
Such an assistant, as all the available evidence suggests, is the one most likely to be controlling the messages, including the message that this is “the most effective means of communication available to me.” And such an assistant, as all the available evidence suggests, may well be the only truly literate person involved in producing such messages.
REFERENCES
Beals, K. (2021) A recent eye-tracking study fails to reveal agency in assisted autistic communication. Evidence-Based Communication Assessment and Intervention, 15:1, 46-51, https://doi.org/10.1080/17489539.2021.1918890
Beals, K. (2024) Can message-passing anecdotes tell us anything about the validity of RPM and S2C? Evidence-Based Communication Assessment and Intervention, https://doi.org/10.1080/17489539.2023.2290298
Gernsbacher M. A. (2004). Language is more than speech: A case study. Journal of Developmental and Learning Disorders, 8, 79–96. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4266410/
Jaswal, V. K., Lampi, A. J., & Stockwell, K. M. (2024). Literacy in nonspeaking autistic people. Autism, 0(0). https://doi.org/10.1177/13623613241230709
Jaswal, V. K., Wayne, A., & Golino, H. (2020). Eye-tracking reveals agency in assisted autistic communication. Scientific reports, 10(1), 7882. https://doi.org/10.1038/s41598-020-64553-9
Leary M. R., Donnellan A. M. (1995). Autism: Sensory-movement differences and diversity. Cambridge Book Review Press.
Torres E. B., Brincker M., Isenhower R. W., Yanovich P., Stigler K. A., Nurnberger J. I., Mextasas D. N., Jose J. V. (2013). Autism: The micromovement perspective. Frontiers in Integrative Neuroscience, 7, Article 32. https://doi.org/10.3389/fnint.2013.00032
Vyse, S. (2020). Of Eye Movements and Autism: The Latest Chapter in a Continuing Controversy. Skeptical Inquirer.
Posts
