perm filename AKIDS[4,KMC]3 blob sn#038873 filedate 1973-05-01 generic text, type T, neo UTF8
00100		THE RATIONALE FOR COMPUTER-BASED TREATMENT OF
00200	        LANGUAGE DIFFICULTIES IN NONSPEAKING AUTISTIC
00300		             CHILDREN
00400	
00500		        KENNETH MARK COLBY
00600	
00700		No one person  could  conceive,  implement  and  utilize  the
00800	treatment method I shall discuss shortly. Hence let me express at the
00900	start my deep indebtedness to my co-workers at  Stanford  University-
01000	Horace  Enea,  David  Smith,  Malcolm Newey and Maxine Colby, each of
01100	whom has put years of effort into this project.
01200	
01300		We  began  about seven years ago with a notion and two facts.
01400	The notion was that a nonspeaking autistic child's primary difficulty
01500	lay  in an inability to process symbols, language being of course the
01600	most important symbolic system used in human communication. The first
01700	fact,  oft  repeated  in the literature, indicated that prognosis was
01800	highly correlated with speech,  the  outlook  for  nonspeakers  being
01900	poor.  The second fact was the common clinical observation that these
02000	children played for hours with machines while  remaining  indifferent
02100	to interactions with people.
02200	
02300		There is now increasing evidence in the  research  literature
02400	supporting   this   notion   of   a   primary  difficulty  in  symbol
02500	processing.[Churchill,1972; Frith,1972; Hermelin  and  O'Connor,1970;
02600	Rutter,  Bartak,  and Newman,1971]. A dysphasic or aphasic child also
02700	has difficulty with  language  but  he  can  acquire  other  symbolic
02800	systems such a gesturing and drawing. He may even learn to read.  But
02900	the nonspeaking autisitc child has great difficulty with all symbolic
03000	processes, not just language.
03100	
03200		The  cause of this difficulty remains unknown.   Nowadays few
03300	experts  in  the  field  defend  a  psychogenic  etiology  since  the
03400	supporting evidence is weak and the disconfirming evidence is gaining
03500	in strength [Rutter,et al,1971]. Regardless of the original cause, if
03600	we  understand  the  crux  of the child's difficulty, we could try to
03700	devise a remedial treatment which takes advantage of  the  fact  that
03800	the  child  is fascinated by machines. Our beginning working idea was
03900	to create a machine a nonspeaking child could play  with in which the
04000	play process involves the child's use of language skills.
04100	
04200		The system we developed consists of a television-like  screen
04300	and  a  typewriter-like  keyboard.  Pressing the keys on the keyboard
04400	causes symbols to appear on the screen accompanied by sounds of human
04500	voices  and  other  noises  common in a child's life. It is much like
04600	having your own Sesame Street to play with. But rather than  being  a
04700	passive  recipient  of  what the television show provides, a child in
04800	our situation is an active initiating agent controlling the  machine.
04900	Instead  of  having  things done to him, things are done by him.  The
05000	merits of the machine are that it is  untiring,  predictable,  always
05100	saying  the  same  thing  the  same way, never angry, never bored and
05200	controllable- properties which are  notoriously  lacking  in  humans.
05300	This  audio-visual-  tactile  experience  is  provided  by a computer
05400	program running on a PDP-6/10 time-sheared computer in  the  Stanford
05500	Artificial  Intelligence  Project.  The program is divided into games
05600	intended to give the child various types of opportunities at  playing
05700	with  and  interacting with symbols. For example, in one game, when a
05800	child presses the key showing the letter  H,  an  H  appears  on  the
05900	screen  and  a  voice  says  "H".  In  another  game pressing the key
06000	labelled "H" produces a running horse on the  screen  accompanied  by
06100	the  sound  of horse's hoofs.  There exist over 1000 such experiences
06200	on  the  system.  The  games  are  organized  at  various  levels  of
06300	complexity  and  are  designed  to  show  a  child how English is put
06400	together from sounds and letters into words and expressions.  I shall
06500	not  go  into  the  details  of  the games here.   They are throughly
06600	described in the literature.[Colby and Smith,1971]. Instead  I  shall
06700	try  to  sketch  the  rationale  or  major principles underlying this
06800	approach.
06900	
07000		First,  consider  how  normal children acquire language. They
07100	are not taught formally as are adults  learning  a  second  language.
07200	Children  are simply exposed to members of a linguistic community and
07300	given  an  opportunity  to  explore  language   usage   in   everyday
07400	communication  between  themselves and other humans who encourage and
07500	correct.  From this exposure,  exploration  and  corrective  feedback
07600	they come to associate the sounds and meanings of words. They come to
07700	sense that certain sounds, human voice sounds, are not just sounds in
07800	themselves  like  the  wind  in  the  trees,  but that they are about
07900	something else, that is, they are  symbolic.   Between  the  ages  of
08000	roughly 2 months and 4 years normal children spend thousands of hours
08100	listening, practicing  and  playing  with  language.   Mastery  of  a
08200	language,  that  is getting most of it right, does not come until far
08300	into adolescence, if ever. A normal child can make an  interpretation
08400	of  expressions  he  has  never  heard before. A nonspeaking autistic
08500	child does not show this course of development.
08600	
08700		The problem is not simply the development of language  skills
08800	but  also  the  acquisition of concepts necessary for a comprehensive
08900	enough model of how the world works, especially the human world.  The
09000	conceptual  or  cognitive  deficits  shown  by  nonspeaking  autistic
09100	children involve those concepts which are normally  acquired  through
09200	language  or  other  symbolizations.    Take  the abstract concept of
09300	`danger'.  To prevent a child  from  becoming  hurt,  a  parent  must
09400	identify  certain  concrete objects and situations as dangerous until
09500	the child grasps the abstract concept of danger.  All  this  is  done
09600	linguistically and through pointing. By way of language , objects and
09700	situations can be referred to and warned about even when they are not
09800	present,  and referred to without pointing when they are present.  An
09900	important socializing function of language is to mark off for a child
10000	what to pay attention to and what is to be done and not done. Many of
10100	the characteristics of autistic children, for example,  the  lack  of
10200	empathy,  can  be  viewed  as  a  consequence of an inability to form
10300	regular conceptual patterns about the  world  because  the  necessary
10400	concepts  acquired  through  language are missing.  As one perceptive
10500	mother said about her autistic child,` a screw is not loose , a screw
10600	is missing'.
10700	
10800		A poorly developed ability to  process  symbols  has  further
10900	consequences  besides  conceptual  deficits. Without language a human
11000	has no awareness of being aware. He lacks the ability to self-monitor
11100	and  to  self-control  by  talking to himself.  He cannot use symbols
11200	reflexively, that is, to to give himself orders  and  to  comment  on
11300	himself  to  himself. Having language a normal child comes to realize
11400	the self is really two which can talk to  one  another.  Finally,  to
11500	become  a  person  one must be recognized as a person by other people
11600	who treat you as if you had  self-awareness  and  self-control.  This
11700	recognition  is  virtually  impossible  for  a child without language
11800	skills and probably represents the greatest tragedy in his life.
11900	
12000		Thus  far  I  have  spoken  of  autistic  children  as if the
12100	nosological label stood for a single homogeneous group. Thirty  years
12200	ago  it  seemed  that  might  be the case. Now it appears as if there
12300	exist several autistic syndromes, none of which  should  be  confused
12400	with  childhood  schizophrenia  since  they  differ in onset, course,
12500	symptoms, family history and  prognosis.    There  are  speaking  and
12600	nonspeaking  autistic children. Among the nonspeaking group there are
12700	those whose linguistic development is normal until  sometime  in  the
12800	second  year when they lose their language abilities. The other major
12900	group are those children who,  from  the  start,  have  trouble  with
13000	language,  understanding  little  and  saying  even less, perhaps one
13100	`mama' or `no' a year. In our experience the most  difficult  problem
13200	for  differential  diagnosis  lies  in deciding whether a nonspeaking
13300	child suffers  from  dysphasia  (developmental  aphasia),  autism  or
13400	perhaps  both.   Over  time the correct diagnosis can be made when it
13500	becomes clear that the dysphasic child can mimic, draw  pictures  and
13600	signify greetings while the autistic child cannot.
13700	
13800	Taking a cue from the normal child who treats language as a toy,  our
13900	first  principle was that the treatment should provide an opportunity
14000	for exploratory play.  The treatment situation is not one  of  forced
14100	drill,  instruction or training but one of play with the keyboard and
14200	video display.  Operant conditioning methods reward  the  child  with
14300	candy  or  food  for  his  actions.   We  do not, believing that food
14400	rewards inhibit exploratory curiosity, as has been shown  in  animals
14500	both  by  Harlow  and  by  Nissen.  Exploratory  learning  requires a
14600	situation which  invites  exploration,  time,  security  and  minimal
14700	interference by adults.
14800	
14900	In each of the sessions the child has a `sitter', an adult whose main
15000	task  is  to  sit  and not interfere. The sitter's task is difficult,
15100	especially if he has been trained to DO things.  We want to give  the
15200	child  an  opportunity  to  freely  self-select  those  symbols which
15300	interest him, rather than to have an adult instruct him or  quiz  him
15400	about  those  symbols which he `should' learn.  The sitter's behavior
15500	is crucial to this treatment method if the spirit of play rather than
15600	drill is to be provided.  Of course the sitter offers social approval
15700	and encouragement when it is fitting.
15800	
15900	An ideal treatment session results when the child is in a good  mood,
16000	is  interested  in working the keyboard display, enjoys imitating the
16100	sounds and is successful in getting the machine to do what he  wants.
16200	The principle of success is important here because, in our view, many
16300	nonspeakers have given up on language.  They  have  failed  over  and
16400	over  and hence withdraw from trying.  We do not let them fail.    As
16500	one normal child said about  the  experience  "it`s  fun,  you  can't
16600	lose". There are students of autism who consider nonspeaking autistic
16700	children to be innately withdrawn from people and hence not acquiring
16800	language.  We  feel  it  is  the  other way around; they have so much
16900	difficulty with language they withdraw from  people  who  unwittingly
17000	flood  and  overwhelm them with meaningless noises. No wonder they do
17100	not call, address, ask  or  answer  questions  of  these  giants  who
17200	immediately  spout  gibberish.   Autistic  children are not aloof and
17300	indifferent to all people -- only to those who  talk.  I  have  often
17400	wondered  if  is not eye-to-eye contact they avoid but eye-to-a-mouth
17500	which , as far as the child is concerned, simply jabbers. If you  say
17600	little  or  nothing  to a nonspeaking child on first meeting him, our
17700	experience has been that you will more than likely find him  in  your
17800	lap, as affectionate as any
17900	other child.
18000	
18100	Another  myth  in  the  literature  which  I  would  like  to dispel.
18200	Autistic children are said to show  pronominal  reversal,  using  the
18300	second  person  singular  pronoun "you" for the first person singular
18400	pronoun "I", and "I" for "you".  This is not so.  It  is  the  normal
18500	child  who  reverses  or  converts  personal pronouns.   The autistic
18600	child LACKS the rule  for  correct  pronoun  assignments  and  simply
18700	echoes  what  he has heard.  Pronouns are examples of deixis in which
18800	the denotation of the words shift depending on the  speech  situation
18900	in  contrast  to,  for example, proper names whose denotations remain
19000	constant. The assignment rule for the variables "I" and "you" is that
19100	the  speaker  calls  himself  "I" and calls the listener "you". Young
19200	autistic children are ignorant of this rule, which  is  not  actively
19300	taught by adults, but is somehow grasped by the normal child from the
19400	data of human dialogues.   The  autistic  child,  unable  to  process
19500	language,  echoes  back,  in  an untransformed and uninterpreted way,
19600	what is said to him.   If you say to him "Do you want your jacket" he
19700	will  echo it, failing to transform the pronouns. Everybody calls him
19800	"you" so he believes "you" must be one of  his  names.   Due  to  the
19900	limitations of short-term memory, an autistic child may echo only the
20000	terminal fragment of a long  expression.  In  English  "I"  regularly
20100	appears  at  the  beginning  of  an expression as the subject. If you
20200	place the pronoun "I" at the end of an expression, an autistic  child
20300	can  echo it as easily as "you". Some people believe that the pronoun
20400	"I" is an index of self-identity and that autistic  children  lack  a
20500	sense of self. From our experience I would say that autistic children
20600	make a very clear distinction between self and non-self. They are  in
20700	fact  hyperautonomous,  as  anyone will find out when he tries to get
20800	the child to do something he does not want to do.  Autistic  children
20900	do  not use "I" and "you" correctly because they are incognizant of a
21000	symbolic transformation rule necessary for the correct assignment and
21100	conversion of input pronouns.  It is worth noting that even in normal
21200	children , "I"  is  never  triggered  by  "you"  but  only  by  "me".
21300	[Fay,1969].
21400	
21500	Returning from this digression on deixis, let me say something  about
21600	our  successes  and  our  failures,  especially  the  latter.   Every
21700	treatment method reports its  dramatic  successes  with  one  or  two
21800	cases.  What  we  need  are  long  case  series before we can compare
21900	methods and decide which is more effective  relative  to  the  effort
22000	required.   Thus  far  our  series  of  nonspeaking autistic children
22100	numbers 17 with 13 of the children improved.  By improvement we mean
22200	only  that  the  child  begins  voluntarily  to use speech for social
22300	communication.  We do not claim the method results in normal language
22400	ability  with  full  comprehension, and correct pronunciation, syntax
22500	and grammar.   Our  aim  is  to  rekindle  the  child's  interest  in
22600	attempting  speech,  to  get  him  to  try  again, to get him off the
22700	ground, to stimulate or catalyze a damaged or slow-developing natural
22800	process   of   language   acquisition.    
22900		We have tried the method  with  other  types  of  nonspeaking
23000	children.  We  failed  with  two  schizophrenic  children.  One brain
23100	damaged child improved and one did not.  We failed completely with  3
23200	children suffering from receptive aphasia. By failure we mean a child
23300	leaves us as he came, perhaps with some  language  understanding  but
23400	producing no useful speech.
23500	
23600	In studying our failures among the autistic group, we have  tried  to
23700	find  some  common  denominator.   We have found only two in our case
23800	series of 16 nonspeaking autistic children. First, all these  failure
23900	cases  are  children  who  showed  no  interest  in  playing with the
24000	machine.  Regardless of our coaxing and persuading  they  would  have
24100	nothing  to  do  with  it.   Finally, becoming desperate, we would be
24200	forced to violate one of our basic principles of allowing free  play.
24300	We  would  spend  weeks and months holding the child at the keyboard,
24400	pushing the keys for him,  trying  to  overcome  his  resistance  and
24500	negativism.   But  to  no avail. Somehow we must be more ingenious in
24600	capturing these children's interest.  We have thought of ways  to  do
24700	this and are currently trying to implement them.
24800	
24900	Second, in cases where the child shows normal linguistic  development
25000	until 16 or 22 months and then suddenly stops talking, we have failed
25100	absolutely.   This history is suggestive of course of  some  type  of
25200	cerebral  insult  such  as a virus infection or auto-immune reaction,
25300	but these hypotheses remain unconfirmed.
25400	
25500	Based on our experience we believe there is  something  powerful  but
25600	not  well-understood  about  the  method.   As  I  described  it,  he
25700	treatment is a complex variable, a shotgun prescription,  the  active
25800	ingredients  of  which are unclear. We need other workers in the field
25900	to adopt the method, improve it and hopefully help us  find  stronger
26000	catalysts for language acquisition.
26100	
26200	Why  haven't  other  workers  tried  this  zero-risk  and potentially
26300	promising method?  Aside from the inertia which all new methods  must
26400	face,  there  are  two  reasons.   One  is  that everyone has his own
26500	preferred method which he believes in and may wish to  improve.   The
26600	other  involves  people's beliefs and misconceptions about computers.
26700	I have come across  people  with  the  following  fantasy  about  our
26800	situation -- that we have a small, cowering, bewildered child sitting
26900	in  front  of  a  towering  bank  of   flashing   lights   {the   New
27000	Yorker-cartoon  concept of a computer} while cold scientists in white
27100	coats observe him thru one-way mirrors.    They  say  "the  child  is
27200	already too interested in machines, he needs relations with people --
27300	you are making him into even more of a robot".   I hope from my brief
27400	description  you  can see this fantasy is not the case.  It may sound
27500	paradoxical, but some nonspeaking children  can  become  more  human,
27600	i.e. become language users, by way of a machine which talks and which
27700	they find more acceptable on their own terms  than  they  do  talking
27800	humans.
27900	
28000	A more realistic objection to computers is their cost.  Few people in
28100	the  world  have  access to million dollar computers for this sort of
28200	work.  But a large computer  is  not  necessary  to  carry  out  this
28300	method.   Mini-computers  costing  only  a  few  thousand dollars are
28400	adequate.  It may be that we do not need a computer at all if someone
28500	found  a  way to rapidly random-accessg both sounds and pictures.  We
28600	are investigating this possibility. Once we can get the  misconceived
28700	Satanic image of the computer out of the picture , I hope others will
28800	join us  in  discovering  more  about  this  technique  for  treating
28900	nonspeaking autistic children.
29000	
29100			REFERENCES
29200	
29300	
29400	[1] Churchill,D.W.,1972. The relation of infantile autism and early
29500	       childhood schizophrenia to developmental language disorders
29600	       of childhood. Journal of Autism and Childhood Schizophrenia,
29700	       2,182-197.
29800	[2] Colby, K.M. and Smith, D.C., 1971. Computers in the treatment of
29900	       nonspeaking autistic children. In Current Psychiatric Therapies,
30000	       Masserman, J.H.(Ed.), Grune and Stratton, New York.
30100	[3] Fay, W.H., 1969. On normal and autistic pronouns. Journal of Speech
30200	       and Hearing,36,242-249.
30300	
30400	
30500	[4] Frith, U., 1972. Cognitive mechanisms in autism: experiments with
30600	        color and tone sequence production. Journal of Autism and
30700	        Childhood Schizophrenia, 2, 160-173.
30800	[5] Hermelin, B. and O'Connor, N., 1970. Psychological Experiments
30900	        With Autistic Children, Pergamon Press, London.
31000	[6] Rutter, M., Bartak, L. and Newman, S., 1971. Autism- a central
31100	        disorder of cognition and language? In Infantile Autism,
31200	        Rutter, M.(Ed.), Churchill Livingstone, London.