00100	.SEC EXPLANATIONS AND MODELS
00200	.SS The Nature of Explanation
00300		It is perhaps as difficult to explain explanation  itself  as
00400	it  is  to  explain  anything  else.     The explanatory practices of
00500	different sciences differ widely but they all share  the  purpose  of
00600	someone   attempting   to   answer   someone   else's  (or  his  own)
00700	why-how-what-etc. questions about a situation, event, episode, object
00800	or phenomenon. Thus explanation implies a dialogue whose participants
00900	share some interests, beliefs, and values.    A consensus must  exist
01000	about   admissable  and  appropriate  questions  and  answers.    The
01100	participants must agree on what is a sound  and  reasonable  question
01200	and  what is a relevant, intelligible, and (believed) correct answer.
01300	The explainer tries to satisfy a  questioner's  curiosity  by  making
01400	comprehensible  why  something  is the way it is. The answer may be a
01500	definition,  an  example,  a  synonym,   a   story,   a   theory,   a
01600	model-description,  etc.  The answer attempts to satisfy curiosity by
01700	settling belief.
01800	.V
01900	    Suppose a man dies and a questioner (Q) asks an explainer (E):
02000	.END CONTINUE
02100	       Q: Why did the man die?
02200	One answer might be:
02300	.V
02400	       E: Because he took cyanide.
02500	.END CONTINUE
02600	This explanation might be sufficient to satisfy Q's curiosity and he
02700	stops asking further questions. Or he might continue:
02800	.V
02900	       Q: Why did the cyanide kill him?
03000	.END CONTINUE
03100	and E replies:
03200	.V
03300	      E: Anyone who ingests cyanide dies.
03400	.END CONTINUE
03500	This explanation appeals to a universal generalization under which is
03600	subsumed  the  particular  fact  of  this  man's  death.  Subsumptive
03700	explanations  satisfy  some  questioners  but  not  others  who,  for
03800	example,  might  want  to  know  about  the  physiological mechanisms
03900	involved.
04000	.V
04100	       Q: How does cyanide work in causing death?
04200	       E: It stops respiration so the person dies from lack of oxygen.
04300	.END CONTINUE
04400		If Q has biochemical interests he might inquire further:
04500	.V
04600		Q:What is  cyanide's  mechanism  of  drug  action  on  the 
04700		  respiratory center?
04800	center?
04900	.END CONTINUE
05000		The last two questions refers to causes. When human action is
05100	to  be  explained,  confusion  easily  arises  between  appealing  to
05200	physical,  mechanical causes and appealing to symbolic-level reasons,
05300	that is, learned, acquired procedures or strategies (Toulmin, 1971).
05400		It  is  established  clinical knowledge that the phenomena of
05500	the paranoid mode can be found associated with a variety of  physical
05600	disorders.    For example, paranoid thinking can be found in patients
05700	with  head   injuries,   hyperthyroidism,   hypothyroidism,   uremia,
05800	pernicious   anemia,   cerebral  arteriosclerosis,  congestive  heart
05900	failure, malaria and epilepsy.      Also drug  intoxications  due  to
06000	alcohol,  amphetamines,  marihuana  and LSD can be accompanied by the
06100	paranoid mode. In these cases the paranoid mode is not a  first-order
06200	disease but a way of processing information in reaction to some other
06300	underlying disorder. To  account  for  the  association  of  paranoid
06400	thought  with  these  physical  states  of  illness,  a psychological
06500	theorist might be tempted to hypothesize that a  purposive  cognitive
06600	system   would  attempt  to  explain  a  physical  illness  state  by
06700	constructing persecutory beliefs blaming other human agents  for  the
06800	ill-being of the disease state. But before making such an explanatory
06900	move, we must consider the elusive distinction  between  reasons  and
07000	causes in explanations of human behavior.
07100		One view  of  the  association  of  the  paranoid  mode  with
07200	physical  disorders  might be that the physical illness simply causes
07300	the paranoia ,through some unknown mechanism, at a  "hardware"  level
07400	beyond the influence of deliberate reprogramming and beyond voluntary
07500	self-control.   That  is,  the  resultant  paranoid  mode  represents
07600	something  that  happens to a person as victim, not something that he
07700	does as an active agent. Another view is that the paranoid  mode  can
07800	be  explained  in  terms of reasons, justifications which describe an
07900	agent's  intentions  and  beliefs.    Does  a  person  as  an   agent
08000	recognize,  monitor  and control what he is doing or trying to do? Or
08100	does  it  just  happen  to  him   automatically   without   conscious
08200	deliberation?   This  question  raises  a  third  view,  namely  that
08300	unrecognized  (but  potentially recognizable) reasons, aspects of the
08400	program which are sealed off and inacessible  to  voluntary  control,
08500	can function like causes.  Once brought to consciousness such reasons
08600	can be modified voluntarily by the agent,  as  a  language  user,  by
08700	reflexively  talking  to  and instructing himself.  This second-order
08800	monitoring and control through language  contrasts  with  an  agent's
08900	inability  to  modify  causes  which  lie  beyond  the  influence  of
09000	self-criticism and self-emancipation through internal  linguistically
09100	mediated  argumentation.    Timeworn  conundrums  about  concepts  of
09200	free-will, determinism, responsibility, consciousness and the  powers
09300	of mental action here plague us unless we stick closely to a computer
09400	analogy which makes a clear and useful distinction between levels  of
09500	hardware, interpreter and programs in a self-referent system. (See p.
09600	000 in Chap 2)
09700	
09800		Each  of these three views provides a serviceable perspective
09900	depending on how a disorder is to be explained and corrected.    When
10000	paranoid  processes  occur during amphetamine intoxication they might
10100	be viewed as biochemically caused and beyond the patient's ability to
10200	control  volitionally through internal self-correcting dialogues with
10300	himself.  When a paranoid moment occurs in a normal person, it can be
10400	viewed  as  having a mistinterpretation as a reason.  If the paranoid
10500	misinterpretation is recognized as such,  a  normal  person  has  the
10600	emancipatory  power  to  revise or reject it through internal debate.
10700	Between these extremes of drug-induced  paranoid  processes  and  the
10800	self-correctible  paranoid moments of the normal person, lie cases of
10900	paranoid personalities  paranoid  reactions  and  the  paranoid  mode
11000	associated    with    the    major   psychoses   (schizophrenic   and
11100	manic-depressive).
11200		One opinion has it that the major psychoses are a consequence
11300	of  unknown  physical  "hardware"  causes  and  are beyond deliberate
11400	voluntary control.  But  what  are  we  to  conclude  about  paranoid
11500	personalities  and  paranoid  reactions where no hardware disorder is
11600	detectable or suspected? Are such persons to be  considered  patients
11700	to  whom something is mechanically happening or are they agents whose
11800	behavior is a consequence of what they do?   Or are they  both  agent
11900	and  patient  depending on on how one views the self-modifiability of
12000	their symbolic processing?  In these perplexing cases we  shall  take
12100	the  position that in normal, neurotic and characterological paranoid
12200	modes, the psychopathlogy represents something that happens to a  man
12300	as  a  consequence  of  what  he  has  experientially  undergone,  of
12400	something he now does, and something he now undergoes.    Thus he  is
12500	both  agent and victim whose symbolic processes have powers to do and
12600	liabilities to undergo.    His liabilities are reflexive in  that  he
12700	is victim to, and can succumb to, his own symbolic structures.
12800	
12900		From this standpoint I  would  postulate  a  duality  between
13000	reasons  and causes. That is, a reason can operate as an unrecognized
13100	cause in one context and be offered as a recognized justification  in
13200	another.    It is, of course, not the reason itself which serves as a
13300	cause  but  having  the  reason.     Human   symbolic   behavior   is
13400	non-determinate  to  the extent that it is self-determinate. Thus the
13500	power to make some decisions freely  and  to  change  one's  mind  is
13600	non-illusory.     When  a reason is recognized to function as a cause
13700	and is accessible to self-monitoring, emancipation from it can  occur
13800	through  change  or rejection of belief. In this sense a two-levelled
13900	system involving an interpreter and its programs  is  self-changeable
14000	and self-emancipatory, within limits.
14100		Explanations  both  in  terms  of  causes  and reasons can be
14200	indefinitely extended and endless questions  can  be  asked  at  each
14300	level of analysis.  Just as the participants in explanatory dialogues
14400	decide what is taken to be problematic, so they  also  determine  the
14500	termini   of   questions   and  answers.   Each  discipline  has  its
14600	characteristic stopping points and boundaries.
14700		In  the  background  of  explanatory dialogues are larger and
14800	smaller constellations of concepts which are  taken  for  granted  as
14900	nonproblematic  background.   Hence  in considering the strategies of
15000	the paranoid mode `it goes without  saying',  that  is,  transcending
15100	this  particular  mode  of  functioning  is  the fact that any living
15200	teleonomic  system  ,as  the  larger  constellation  ,  strives   for
15300	maintenance  and  expansion of life. Also it should go without saying
15400	that, at a lower level, ion transport takes place through  nerve-cell
15500	membranes.  Every function of an organism can be viewed a governing a
15600	subfunction beneath and depending  on  a  transfunction  above  which
15700	calls it into play for a purpose.
15800		Just as there are many alternative ways of describing,  there
15900	are many alternative ways of explaining.  An explanation is geared to
16000	some  level  of  what  the  dialogue  participants  take  to  be  the
16100	fundamental  structures  and processes under consideration.  Since in
16200	psychiatry   we   cope   with   patients'   problems   using   mainly
16300	symbolic-conceptual  techniques,(although  it  is true that the pill,
16400	the knife, and electricity are also available.), we are interested in
16500	aspects  of  human  conduct  which  can be explained, understood, and
16600	modified at a symbol-processing level. Psychiatrists need theoretical
16700	symbolic   systems  from  which  their  clinical  experience  can  be
16800	logically derived to interpret the case histories of their  patients.
16900	Otherwise  they  are  faced  with mountains of dross and indigestible
17000	data.   "Science is an attempt to make the chaotic diversity  of  our
17100	sense  experience correspond to a logically uniform system of thought
17200	by correlating single experiences  with  the  theoretic  structure."-
17300	Einstein.
17400	
17500	.SS The Symbol Processing Viewpoint
17600	
17700		Segments  and  sequences  of  human behavior can be looked at
17800	from many standpoints.   In this monograph I shall view sequences  of
17900	paranoid symbolic behavior from an information processing standpoint.
18000	For  a  more  complete  explication   and   justification   of   this
18100	symbol-processing view, see Newell (1973) and Newell and Simon (1972).
18200		In brief, information is defined as knowledge in  a  symbolic
18300	code.    A symbolic process is a symbol-manipulating activity posited
18400	to account  for  observable  symbolic  behavior  such  as  linguistic
18500	interaction.  Symbols  are  defined  as representations of experience
18600	classified as objects, events, situations, and  relations.
18700		Symbol-processing  explanations   postulate   an   underlying
18800	structure   of  hypothetical  processes,  functions,  strategies,  or
18900	directed symbol-manipulating procedures, having the power to  produce
19000	and  being  responsible  for the manifest phenomena. Such a structure
19100	offers  an  ethogenic  (ethos  =  conduct  or  character,   genic   =
19200	generating)   explanation  for  sequences  or  segments  of  symbolic
19300	behavior. (See Harre  and  Secord,1972).  In  adopting  an  ethogenic
19400	viewpoint,   I  shall  posit  processes,  functions,  procedures  and
19500	strategies as being responsible for and having the power to  generate
19600	the  symbolic  patterns  and sequences characteristic of the paranoid
19700	mode.   "Strategies" is  perhaps  the  best  general  term  since  it
19800	implies  ways  of  obtaining  an  objective which have suppleness and
19900	pliability   since   their   choice   of   application   depends   on
20000	circumstances.        However   I   shall   use   all   these   terms
20100	interchangeably.
20200	
20300	.SS Symbolic Models
20400		Theories and  models  share  many  functions  and  are  often
20500	considered  equivalent.  One  important  distinction lies in the fact
20600	that a theory states a subject has a certain structure but  does  not
20700	exhibit  that  structure in itself. (See Kaplan,1964). In the case of
20800	interactive simulation models,  such  as  will  be  described,  there
20900	exists  a  further  distinction.      Interactive  simulation models,
21000	having the ability to converse in natural language  using  teletypes,
21100	actualize or realize a theory in the form of a dialogue algorithm. In
21200	contrast to a verbal, pictorial or mathematical representation,  such
21300	a model changes its states over time and ends up in a state different
21400	from its initial state.
21500		In  contrasting  description from what is described, Einstein
21600	remarked  that it is not the function of science to give the taste of
21700	the soup.  But an interactive simulation  model  which  reproduces  a
21800	segment  of  reality does just that, since it offers an interviewer a
21900	first-hand experience  with  a  concrete  case.   In  constructing  a
22000	computer  simulation, a theory is modelled to discover a sufficiently
22100	rich structure of assumptions to generate the observable behavior
22200	under  study.    A  dialogue algorithm allows an observer to interact
22300	with a concrete specimen of a class in detail. In  the  case  of  our
22400	model,  the  level of detail is the level of the symbolic behavior of
22500	conversational language which is satisfying to a  clinician  who  can
22600	compare  the  model  with human counterparts at his familiar level of
22700	clinical dialogue. Communicating with the paranoid model by means  of
22800	teletype, an interviewer can directly experience for himself the type
22900	of impaired  social  relationship  which  develops  with  someone  in
23000	paranoid mode.
23100		An algorithm composed of  symbolic  computational  procedures
23200	converts  input  symbolic  structures into output symbolic structures
23300	according to certain principles.  The modus operandi  of  a  symbolic
23400	model  is simply the workings of an algorithm when run on a computer.
23500	At this level of explanation, to answer `why?' means  to  provide  an
23600	algorithm  which  makes explicit how symbolic structures collaborate,
23700	interplay and interlock  -  in  short,  how  they  are  organized  to
23800	generate patterns of manifest phenomena.
23900	
24000		To  simulate the sequential input-output behavior of a system
24100	using symbolic  computational  procedures,  we  write  an  alogorithm
24200	which,  when run on a computer, produces symbolic behavior resembling
24300	that  of  the  subject  system  being  simulated.   (Colby,1973)  The
24400	resemblance  is  achieved  through  the  workings of an inner posited
24500	structure  in  the  form  of  an  algorithm,   an   organization   of
24600	symbol-manipulating   procedures   which   are  responsible  for  the
24700	characteristic observable behavior at the input-output level.   Since
24800	we  do not know the structure of the `real' simulative processes used
24900	by the mind-brain,  our  posited  structure  stands  as  an  imagined
25000	theoretical  analogue,  a  possible  and  plausible  organization  of
25100	processes analogous to  the  unknown  processes  and  serving  as  an
25200	attempt  to  explain  the  workings  of  the  system  under study.  A
25300	simulation model is thus deeper than  a  pure  black-box  explanation
25400	because  it  postulates  functionally equivalent processes inside the
25500	box to account for observable patterns  of  behavior.   A  simulation
25600	model  constitutes  an  interpretive  explanation  in  that  it makes
25700	intelligible the connections between external input, internal  states
25800	and  output  by  positing  intervening  symbol-processing  procedures
25900	operating  between  symbolic  input  and   symbolic   output.      An
26000	intelligible  description  of the model should make clear why and how
26100	it reacts as it does under various circumstances.
26200		Citing a universal generalization to explain an  individual's
26300	behavior  is unsatisfactory to a questioner who is interested in what
26400	powers and liabilities are latent behind manifest phenomena.  To  say
26500	`x is nasty because x is paranoid and all paranoids are nasty' may be
26600	relevant, intelligible and correct. But another type  of  explanation
26700	is  possible,  a model-explanation referring to a structure which can
26800	account for `nasty' behavior as a consequence of input  and  internal
26900	states  of  a  system.   A  model  explanation  specifies  particular
27000	antecedants  and  processes  through  which  antecedants generate the
27100	phenomena.   An ethogenic approach to explanation assumes perceptible
27200	phenomena  display the regularities and nonrandom irregularities they
27300	do  because  of  the  nature  of  a  imperceptible  and  inaccessible
27400	underlying  structure.   The  posited  theoretical  structure  is  an
27500	idealization, unobservable in human heads,  not  because  it  is  too
27600	small, but because it is imaginary.
27700		When attempts are made to explain human behavior,  principles
27800	in  addition  to  those accounting for the natural order are invoked.
27900	"Nature entertains no opinions about us", said Nietzsche,  but  human
28000	natures  do  ,  and  therein  lies  a  source  of  complexity for the
28100	understanding of human conduct. Until the first quarter of  the  20th
28200	century,  natural sciences have been guided by the Newtonian ideal of
28300	perfect process knowledge about inanimate objects whose behavior  can
28400	be  subsumed  under lawlike generalizations.  When a deviation from a
28500	law  was  noticed,it  was  the  law  which  was  modified,  since  by
28600	definition physical objects do not have the power to break laws. When
28700	the planet Mercury was observed to deviate from the  orbit  predicted
28800	by  Newtonian  theory,  no  one  accused  the  planet  of  being   an
28900	intentional agent breaking the law; something was incorrect about the
29000	theory.    Subsumptive explanation is the acceptable norm in  physics
29100	but  it  is  seldom  satisfactory  in  accounting for the behavior of
29200	living purposive systems.     In considering the behavior  of  bodies
29300	falling   in  a  macroscopic  world,  no  one  nowadays  follows  the
29400	Aristotelian pattern of attributing to them intentions to fall .  But
29500	in  the  case  of  living  systems,  especially  ourselves, our ideal
29600	explanatory practice remains Aristotelian in utilizing a  concept  of
29700	intention.  Aristotle's misconception in physics was to extend to the
29800	macroscopic non-living world an intentionalistic concept  of  purpose
29900	appropriate  to  the  living world as a principle of intelligibility.
30000	(See Ayala,1972).
30100		Consider  a  man participating in a high-diving contest.   In
30200	falling towards the water he accelerates at the rate of 32  feet  per
30300	second. Viewing the man simply as a falling body, we explain his rate
30400	of fall by appealing to a physical law.  Viewing the man as  a  human
30500	intentionalistic  agent,  we  explain  his  dive  as the result of an
30600	intention to dive in a cetain way in order to win the diving contest.
30700	His  conduct  (in  contrast  to  mere  movement) involves an intended
30800	following of certain conventional rules for what is judged by  humans
30900	to  constitute, say, a swan dive. Suppose part way down he chooses to
31000	change his position in mid-air and enter the water thumbing his  nose
31100	at  the  judges. He cannot break the law of falling bodies but he can
31200	break the  rules  of  diving  and  make  a  gesture  which  expresses
31300	disrespect  and  which he believes will be interpreted as such by the
31400	onlookers.   Our diver breaks a rule for diving but  follows  another
31500	rule  which  prescribes  gestural action for insulting behavior.   To
31600	explain the actions of diving and nose-thumbing, we would appeal, not
31700	to  laws  of natural order, but to an additional order, to principles
31800	of human order, superimposed on laws of natural order and which  take
31900	into account (1)standards of appropriate action in certain situations
32000	and (2) the agent's inner considerations  of  intention,  belief  and
32100	value  which he finds compelling from his point of view. In this type
32200	of explanation the explanandum, that which is being explained is  the
32300	agent's  informed  actions,  not  simply his movements.  When a human
32400	agent performs an action in a situation, we can ask:  is  the  action
32500	appropriate  to  that situation and if not, why did the agent believe
32600	his action to be called for.
32700		As  will  be  shown,  symbol-processing  explanations rely on
32800	concepts of intention, belief, action, affect, etc. These  terms  are
32900	close to the terms of ordinary language as is characteristic of early
33000	stages of explanations. It is also important to note that such  terms
33100	are  commonly utilized in describing computer algorithms which strive
33200	to achieve goals.  In  an  algorithm  these  ordinary  terms  can  be
33300	explicitly defined and represented.
33400		Psychiatry deals with the practical concerns of inappropriate
33500	action, belief, etc. on the part of a patient. His  behavior  may  be
33600	inappropriate  to  the  onlooker since it represents a lapse from the
33700	expected, a contravention of the human order. It may even appear this
33800	way  to  the  patient  in  monitoring  and  directing  himself.   But
33900	sometimes, as in severe cases of  the paranoid  mode,  the  patient's
34000	behavior  does  not  appear  anomalous to himself.  He maintains that
34100	anyone  who  understands  his  point  of  view,  who   conceptualizes
34200	situations  as  he  does  from the inside, would consider his outward
34300	behavior appropriate and justified. What he does  not  understand  or
34400	accept is that his inner conceptualization is mistaken and represents
34500	a misinterpretation of the events of his experience.
34600		The  model  to  be  presented  in  the  sequel constitutes an
34700	attempt to explain some regularities and  particular  occurrences  of
34800	symbolic   (conversational)   paranoid  behavior  observable  in  the
34900	clinical situation of a psychiatric interview.   The  explanation  is
35000	at the symbol-processing level of linguistically communicating agents
35100	and  is  cast  in  the  form  of  a  dialogue  algorithm.  Like   all
35200	explanations  it  is  incomplete  and does not claim to represent the
35300	only conceivable structure of processes .
35400	
35500	The Nature of Algorithms
35600	
35700		Theories can be presented in various forms  such  as  essays,
35800	mathematical   equations   and   computer   programs.  To  date  most
35900	theoretical explanations in psychiatry and psychology have  consisted
36000	of  natural  language  essays with all their well-known vagueness and
36100	ambiguities. Many of these formulations  have  been  untestable,  not
36200	because relevant observations were lacking but because it was unclear
36300	what the essay was really saying. Clarity is needed.
36400		An  alternative  way of formulating psychological theories is
36500	now available in the form of symbol-processing  algorithms,  computer
36600	programs,  which have the virtue of being clear and explicit in their
36700	articulation and which can be run on  a  computer  to  test  internal
36800	consistency and external correspondence with the data of observation.
36900	The subject of a model is what it is a model  of;  the  source  of  a
37000	model  is  what  it  is  based  upon. Since we do not know the `real'
37100	mind-brain algorithms, we construct a theoretical model,  bades  upon
37200	computer algorithms, which  represents  a  partial  analogy.  (Harre,
37300	1970).   The  analogy is made at the symbol- processing level, not at
37400	the hardware  level.    A  functional,  computational  or  procedural
37500	equivalence  is  being postulated.   The question then becomes one of
37600	categorizing  the  extent  of  the  equivalence.        A   beginning
37700	(first-approximation)  functional  equivalence  might  be  defined as
37800	indistinguishability at the  level  of  observable  I-O  pairs.     A
37900	stronger  equivalence  would consist of indistinguishability at inner
38000	I-O levels. That is, there exists a correspondence  between  what  is
38100	being done and how it is being done at a given level of operations.
38200		An algorithm represents an organization of  symbol-processing
38300	strategies  or functions which represent an `effective procedure'. It
38400	is  essential  to  grasp  this  fundamental   concept   of   computer
38500	simulation. An effective procedure consists of two compoments:
38600	.V
38700		(1) A programming language in which procedural rules of
38800		    behavior can be rigorously and unambiguously specified.
38900	
39000		(2) A machine processor which can rapidly and reliably carry
39100		    out the processes specified by the procedural rules.
39200	.END
39300	The specifications of (1), written in a formally defined  programming
39400	language,  is  termed  an  algorithm  or program while (2) involves a
39500	computer as the machine processor, a set  of  deterministic  physical
39600	mechanisms   which  can  perform  the  operations  specified  in  the
39700	algorithm. The algorithm is called `effective'  because  it  actually
39800	works, performing as intended when run on the machine processor.
39900		A simulation model is composed of procedures analogous to the
40000	real   and  unknown  procedures.    We  are  not  claiming  they  ARE
40100	analogous, we are MAKING them so. The analogy  being  drawn  here  is
40200	between specified processes and their generating systems. Thus
40300	
40400	.V
40500	      mental process    computational process
40600	      --------------:: ----------------------
40700	      brain hardware      computer hardware and
40800	      and programs           programs
40900	.END
41000	
41100		Many of the classiclal mind-brain problems arose  because  we
41200	did  not yet have for analogy a familiar example of a system in which
41300	we could make a clear separation between  hardware  descriptions  and
41400	program  descriptions. With the advent of computers and programs some
41500	mind-brain perplexities disappear.  (Colby,1971).  The analogy is not
41600	simply  between  computer  hardware  and  brain  wetware.  We are not
41700	comparing  the  structure  of   neurons   with   the   structure   of
41800	transisitors;  we are comparing the organization of symbol-processing
41900	procedures in an algorithm with symbol-processing procedures  of  the
42000	mind-brain.  The  central nervous system contains a representation of
42100	the experience of its holder.   A  model  builder  has  a  conceptual
42200	representation  of  that  representation which he demonstrates in the
42300	form  of  a  model.  Thus  the  model  is  a   demonstration   of   a
42400	representation of a representation.
42500		Since we are taking running computer programs as a source  of
42600	analogy  for a paranoid model, errors or pathological behavior on the
42700	part of such programs are of interest to the psychopathologist. These
42800	errors  can  be ascribed to the hardware level, to the interpreter or
42900	to the programs which the interpreter  executes.  Different  remedies
43000	are  required  at different levels. If the analogy is to be useful in
43100	the case of human pathological behavior, it will become a  matter  of
43200	influencing symbolic behavior with the appropriate techniques.
43300		Since the algoritm is written in a programming  language,  it
43400	is  hermetic  except  to  a  few  people, who in general do not enjoy
43500	reading  other  people's  code.    Hence  the   intelligibility   and
43600	scrutability  requirement for explanations must be met in other ways.
43700	In an attempt to open the model to  scrutiny  I  shall  describe  the
43800	model in detail using diagrams and interview examples profusely.
43900	
44000	
44100	Analogy
44200	
44300		I  have  stated  that  an  interactive  simulation  model  of
44400	symbol-manipulating   processes   reproduces  sequences  of  symbolic
44500	behavior at the level of linguistic communication.  The  reproduction
44600	is  achieved  through the operations of an algorithm consisting of an
44700	organization  of   hypothetical   symbol-processing   strategies   or
44800	procedures  which  can  generate  the  I-O  behavior  of the subject-
44900	processes under  investigation.The  algorithm  is  be  an  "effective
45000	procedure" in the sense it really works in the manner intended by the
45100	model-builders. In the model to be described, the paranoid  algorithm
45200	generates   linguistic   I-O   behavior  typical  of  patients  whose
45300	symbol-processing is dominated by the paranoid mode. Comparisons  can
45400	be  made  between samples of the I-O behaviors of patients and model.
45500	But the analogy is not to be drawn at this level.   Mynah  birds  and
45600	tape  recorders  also  reproduce human linguistic behavior but no one
45700	believes the reproduction is achieved by powers  analogous  to  human
45800	powers.   Given that the manifest outermost I-O behavior of the model
45900	is indistinguishable  from  the  manifest  outward  I/O  behavior  of
46000	paranoid  patients,  does this imply that the hypothetical underlying
46100	processes used by the model are analogous  to  or  the  same  as  the
46200	underlying processes used by persons in the paranoid mode?  This deep
46300	and far-reaching question should be approached with caution and  only
46400	when  we  are  first  armed  with  some  clear notions about analogy,
46500	similarity,   faithful   reproduction,    indistinguishability    and
46600	functional equivalence.
46700		In comparing two things (objects, systems or processes )  one
46800	can   cite   properties   they  have  in  common,(positive  analogy),
46900	properties they do not share (negative analogy) and properties  which
47000	we  do  not  yet  know whether they are positive or negative (neutral
47100	analogy). (See Hesse,1966). No two things are exactly alike in  every
47200	detail.   If  they  were identical in respect to all their properties
47300	then they would be copies. If they were identical  in  every  respect
47400	including  their  spatio-temporal  location we would say we have only
47500	one thing instead of two. Everything  resembles  something  else  and
47600	maybe everything else, depending upon how one cites properties.
47700		In an analogy a similarity relation is  evoked.  "Newton  did
47800	not  show  the  cause of the apple falling but he showed a similitude
47900	between the apple and the stars."(D`Arcy Thompson). Huygens suggested
48000	an analogy between sound waves and light waves in order to understand
48100	something less well-understood (light) in terms of  something  better
48200	understood   (sound).   To  account  for  species  variation,  Darwin
48300	postulated a  process  of  natural  selection.    He  constructed  an
48400	analogy  from two sources, one from artificial selection as practiced
48500	by domestic breeders of animals and one from  Malthus'  theory  of  a
48600	competition  for  existence  in a population increasing geometrically
48700	while its resources increase arithmetically. Bohr's model of the atom
48800	offered  an  analogy  between solar system and atom. These well-known
48900	historical examples should be sufficient here to illustrate the  role
49000	of analogies in theory construction.    Analogies are made in respect
49100	to  those  properties  which  constitute  the  positive  and  neutral
49200	analogy.     The  negative analogy is ignored.   Thus Bohr's model of
49300	the atom as a miniature planetary system was not intended to  suggest
49400	that  electrons  possessed  color or that planets jumped out of their
49500	orbits. 
49600	
49700	Functional Equivalence
49800	
49900		When human symbolic processes are the subject of a simulation
50000	model, we draw from two sources, symbolic computation and psychology.
50100	We propose an analogy between systems known  to  have  the  power  to
50200	process  symbols,  namely, persons and computers.      The properties
50300	compared in the analogy are obviously  not  physical  or  substantive
50400	such  as blood and wires, but functional and procedural.   We want to
50500	assume that the not-well-understood procedures of thought in a person
50600	are  similar  to the more accessible and better understood procedures
50700	of symbol-processing which take place in a computer.   The analogy is
50800	one  of  functional or procedural equivalence. (For a further account
50900	of  functional  analysis  see  Hempel   (1965)).     Mousetraps   are
51000	functionally  equivalent.    There  exists  a  large  set of physical
51100	mechanisms for catching mice. The term "mousetrap" says what  all  of
51200	the set has in common.   They take as input a live mouse and yield as
51300	output a dead one. Systems equivalent from one point of view may  not
51400	be equivalent from another. (Fodor,1968).
51500		If model and human  are  indistinguishable  at  the  manifest
51600	level of linguistic I-O pairs, then they can be considered equivalent
51700	at that level.     If they can be shown to  be  indistinguishable  at
51800	more  internal  symbolic  levels, then a stronger equivalence becomes
51900	achieved.    How stringent and how  extensive  are  the  demands  for
52000	equivalence to be?    Must there be point-to-point correspondences at
52100	every level?   What is to count as a point and what are  the  levels?
52200	Procedures  can  be  specified  and  ostensively  pointed  to  in  an
52300	algorithm but how can we point to unobservable symbolic processes  in
52400	a  person's  head?   There is an inevitable limit to scrutinizing the
52500	"underlying" processes of the world.  Einstein likened this situation
52600	to  a  man  explaining  the  behavior  of a watch without opening it:
52700	"He will never be able to compare his picture with the real mechanism
52800	and  he  cannot  even  imagine  the  possibility or meaning of such a
52900	comparison". 
53000		In  constructing  an   algorithm   one   puts   together   an
53100	organization  of  collaborating  functions or procedures.  A function
53200	takes some symbolic structure  as  input  and  yields  some  symbolic
53300	structure as output. Two computationally equivalent functions, having
53400	the same input and yielding the same output, can differ `inside'  the
53500	function at the instruction level.
53600		Consider  an elementary programming problem which students in
53700	symbolic computation are often asked to solve.  Given  a  list  L  of
53800	symbols,  L=(A  B  C  D), as input, construct a function or procedure
53900	which will convert this list to the list RL in which the order of the
54000	symbols  is  reversed,  i.e.   RL=(D  C B A).  There are many ways of
54100	solving this problem and the code of one student may  differ  greatly
54200	from that of another at the level of individual instructions. But the
54300	differences of such details are irrelevant. What  is  significant  is
54400	that  the  solutions  make  the required conversion from L to RL. The
54500	correct solutions will  all  be  computationally  equivalent  at  the
54600	input-output  level  since  they take the same symbolic structures as
54700	input and produce the same symbolic output.
54800		If  we  propose  that  an  algorithm  we  have constructed is
54900	functionally equivalent to what goes on in humans when  they  process
55000	symbolic   structures,   how   can   we   justify   this  position  ?
55100	Indistinguishability tests at,  say,  the  linguistic  level  provide
55200	evidence  only for beginning equivalence. We would like to be able to
55300	have access to the underlying processes in humans the way we can with
55400	algorithms.  (Admittedly,  we  do  not directly observe processes but
55500	only their products). The  difficulty  lies  in  identifying,  making
55600	accessible,   and   counting   processes   in   human  heads.    Many
55700	symbol-processing experiments are now being designed and carried out.
55800	We   must   have  great  patience  with  this  type  of  experimental
55900	psychology.
56000		In  the meantime, besides first-approximation I-O equivalence
56100	and plausibility arguments,  one  might  appeal  to  extra-evidential
56200	support  offering parallelisms from other scientific domains. One can
56300	offer analogies between what is known to go on at a  molecular  level
56400	in  the  cells  of living organisms and what goes on in an algorithm.
56500	For example, a DNA molecule in the nucleus of a cell consists  of  an
56600	ordered  sequence  (list)  of  nucleotide  bases  (symbols)  coded in
56700	triplets termed codons (words). Each element of the  codon  specifies
56800	which  amino  acid  during protein synthesis is to be linked into the
56900	chain of polypeptides making up the  protein.   The  codons  function
57000	like instructions in a programming language. Some codons are known to
57100	operate as terminal symbols analogous  to  symbols  in  an  algorithm
57200	which  terminate  the end of a list. If, as a result of a mutation, a
57300	stop codon should appear in the middle of a sequence rather  than  at
57400	its  normal  terminal  position, further protein synthesis is halted.
57500	The polypeptide chain resulting is abnormal and may  have  lethal  or
57600	trivial  consequences  for  the  organism  depending  on what must be
57700	passed on to other processes which require polypeptides to be  handed
57800	over  to  them. Similarly in an algorithm. If a terminating symbol is
57900	incorrect in  a  procedure,  the  procedure  cannot  operate  in  its
58000	intended  manner.  Such  a  result  may  be  lethal or trivial to the
58100	algorithm depending on what information  the  faulty  procedure  must
58200	pass  on  at  it  interface  with  other  procedures  in  the overall
58300	organization. Each procedure  in  an  algorithm  is  embedded  in  an
58400	organization  of  collaborating  procedures  just  as  is the case of
58500	functions in living organisms. We know that at the molecular level of
58600	living  organisms  there  exists a process such as serial progression
58700	along a nucleotide sequence, which is analogous to  stepping  down  a
58800	list  in  an  algorithm.  Further analogies can be made between point
58900	mutations in which DNA codons can be inserted,  deleted,  substituted
59000	or  reordered  and  symbolic computation in which the same operations
59100	are commonly carried out on symbolic structures.   Such analogies are
59200	interesting as extra-evidential support but obviously closer linkages
59300	are needed between the macro-level  of  symbolic  processes  and  the
59400	micro-level of molecular information-processing within cells.
59500		To  obtain  evidence  for  the  acceptability  of  a   model,
59600	empirical  tests  are utilized as validation procedures.   Such tests
59700	should also tell us which is the best among alternative versions of a
59800	family of models and among alternative,different families of, models.
59900	Scientific explanations do not stand alone in isolation.    They  are
60000	evaluated  relative  to  rival  contenders  for the position of "best
60100	available". Once we accept a theory or model as the  best  available,
60200	can  we  be  sure  it  is  correct or true?    We can never know with
60300	certainty. Theories and  models  are  provisional  approximations  to
60400	nature destined to become superseded by better ones.