LETTERS

ective Testing Lateralization

Proj

and

ever

have

to work

with.

It is prob-

tive, sonal,

the

context-dependent

(con-

SIR: I found tive Bias Task

tal Lobe

the

article

The

Cogni-

ably not accurate at a neural network level to think in terms of behavioral outputs as stemming from complex peting, parallel In this single representations of adap-

ventional?)
more

response
guarded,

is less perless and freely paralimbic empirical

and

Lateralized
in Males1

Fronto be

and by right-hemisphere

associative, activations.

thus

less mediated

Functions by the

tive context,

as opposed

to being
comand

Given

much

most
nated cepts digms

interesting,

having been fasciconfusing literature on

for years. The conand as parafrontal procresearched I only

amalgams of multiple variably interdigitating, representations.

lateralization context-independent

of context-dependent for left and right are compellingly very well presented.

pendent sensory
that and termed sonal posed is curious pened

sense, the context-inderesponses seem to reflect biases (likes and dislikes)

have important limbic

evidence that right-hemisphere activation is more intrinsically and critically tied into limbic/paralimbic activation, this potential connection of the Cognitive jective testing ing empirically

Bias Task

with

pro-

essing
and

probably paralimbic

could be worth pursuin a much more

wish

that

the article
published

had
work

addressed
relevant lobes.

der if they

aspects, and I wonmight be more accurately

idiosyncratically peras opIt have very hapuse-

systematic
A gulf literatures sentation

way.
continues to separate two that have looked at reprefrom disparate, almost on the one and other im-

additional to lateralization

more

of the

frontal

There were also some inherently problematic concepts concerned with the nature of executive func-

or less constrained to context-independent. that upon the authors a potentially personal

miscible, perspectives: hand, psychoanalysis

psychodynamic
on the
grounded

psychologies,

and,

tions,
there sponses.

such
are Im

as the suggestion
context-invariant not really sure

that
rewhat and the

ful paradigm,

bias
long

to ambiguous
history connection to the

tasks,

response that has a

without

other,

psychologies,
methodologies ioral bent.

in psychiatry, having authors. This

such
than like.

responses
deep I doubt exceptions, tendon that

would
complex

be, other
behavior,

that
parent

become
para-

ap-

reflexes

various cognitive especially those in cognitive psychometric with a strong behavOne perspective tends to

digm-projective

testing-is

assume
tant (mostly) spective

that

all of the really

schemes

imporare this perobtuse object (what now poten-

outside
other stripped

of ictal

events
can

and
ever

a few
be to matrices

of a connection

of internal representations, including representations of adaptive context. The questions nature are about the reprewould activato be. differential of those

sentations in each hemispherewhat neural network theory term the tion-and types The state how space these of their differential come

of constructions notion

of a context-invari-

ant
tion that

response

may

be a contradicevidence mediate I mean context, those context fully consciously coded in lanre-

in terms, given much context must critically By context or represented here

all output.
perceived

gardless of whether representations are available guage. authors context real be) text, context instead which or potentially

There are times when the appear to be using the term as though (whatever of the in fact it means that is all that represented we the might con-

generally assumed, erroneously in my judgment, to have nothing to offer clinical neuroscience (P. Malloy, personal communication on assessing Capgras patients by Rorschach versus other projective tests, 1992). But this work on the Cognitive Bias Task suggests that projective tests should get a second look in terms of their functional value as probes of lateralized or regions-of-interest activation. My own thinking has always been that tests such as the Rorschach and the Thematic Apperception Test tap paralimbic and right-hemisphere representational systems-but only when the person moves away from the popular or conventional safe, more contextdependent response. Whole scoring systems in dunical psychology are in fact based on this distinction between conventional and idiosyncratic responses. This hypothesis is entirely consistent with the authors findings. Perhaps from this perspec-

representational

unconscious.
we have

From
relatively

notions
relations neural

such
network

as internalized
theory would of the

in psychoanalysis

call nonverbal maps tial interpersonal-affective space). sumes of the that above The other the opposite really important drive the behavior water

state
asall notion, that cognitions are

perspective

conscious,
and fairly

line,

easily coded in language (at least potentially). Obviously, when stated in such terms, both assumptions look
shaky. a serious of the ing complex I do think, mistake critical however, to assume that that guidespecially issues, line of it is all

representations behavior,

when dealing with affective can be seen above the water consciousness. trained Certainly in exploratory would believe article only two

no one psychother-

apy ences

that.
refermany dozens

The

unfortunately of the

232

VOLUME

#{149} NUMBER

#{149} SPRING

1996

LETTERS

of major eralization berg and essing

theoretical overviews of latof the brain, one by GoldCosta, outlining evidence versus novel proc8.

fundamental

aspects

of attention
tied to the right the neurodefor this almost
3.

connection
by Dimond

and Cerebral
SJ, Beaumont
1974, pp

Function,

edited

being intrinsically hemisphere, with velopmental totally basis unknown.7

JG. London,
167-183

Elek Science,

of a routinized

distinction, another by Tucker and Williamson, outlining neuromodulatory differences, and

The neuromodulatory noted in the authors

view

differences literature reand seroto-

the authors
own ideas viewpoint in these

do not integrate
on laterality other major

their
with reviews. even if prefrontal to address ideas and state the

(norepinephrine

4.

Watt DF: Higher cortical functions and the ego: explorations of the boundary between behavioral neurology, neuropsychology, and psychoanalysis. Psychoanalytic Psychology 1990; 7:487527 Goldberg E, Costa LD: Hemispheric differences in the acquisition and use of de-

nm more right, dopamine more left); neuroanatomical differences that

netal

scriptive systems. 14:144-173

5. Joseph

Brain Lang 1981;

NeuropsyNeurology. New

Reviews they are functions, heuristically findings

of lateralization, only examining probably need several about basic laterality. must on

were

not

noted

(larger larger

left right pa-

R: Neuropsychology, and Behavioral

chiatry,
6.

planum among

temporale, and prefrontal others); and

regions the possibility differences, of intact patients, of the two tha-

York, Plenum, 1990 Galin D: Implications left and right cerebral

Arch Gen Psychiatry

7.

for psychiatry of specialization. 1974; 31:572-583

lim-

Or if this
put for 9.

of cytoarchitectonic

synthesis
of the following eventual 1. art,

is beyond
one pieces synthesis:

the current
at least the table

cortically and subcortically. The curious relative absence conscious and evidence division that either in both commissurotomized

Mesulam M-M: Patterns in behavioral neuroanatomy: association areas, the bic system, and hemispheric specialization, in Principles of Behavioral Neurology, edited by Mesulam M-M. Philadelphia, FA Davis, 1985, pp 1-70

Complementation analyzer Digital types

of an opposed

lamocortical
to sustain ness and

systems
higher a sense

appears

able

syntactical-phonological-semantic 2. and a Gestalt synthesizer.2 versus analogue encoding paradigms that must reflect different
Routinized

order consciousof self, however

In Reply
SIR:
Watt nent, issues: The letter in response by Dr. Douglas to our recent F. arti-

different they may be in affective tone and modal style of processing. This any tive is clearly a vast to digest, literature but for of execuit merits

3.

of recurrent (left)

cortical versus

networks.3

cle touches
review functions of lateralization

on a number

novel

4.

(right) processing loads.4 Differential affective valences-the

right

but not directly the projective the notion the neural

of pertiinterrelated, nature of
in-

more euphoric

dysphoric, integration

the left of the af-

more 5. fective essing two

in valence. issues and the procissues, which are often approached re-

attention. One neural network tion will emerge

hopes that artificial models of lateralizafrom artificial neuwork that all of these fea-

our task; variance; the multitude

of context net models

of hemispheric

specialization;
of theories We Task will

and
of hemiad-

A theoretical valence modality questions

ral network-connectionist may help integrate tures into one But that seems pect,
current

quite separately with little good sult. Why is the Gestalt synthesizer/analogical processor/novel eventlcontext-independent essor more involved tion of negative affect, digitallroutinized/semantically while

heuristic explanation. not to be a near pros-

spheric specialization. dress them in turn. The Cognitive Bias tive must in the sense disambiguate task

is projec-

since

I know

of not one

single

that the subject an inherently by introducing his

procthe

tion, several

ANN despite leading

model for lateralizadiscussion of this with neural personal Pi-i.D. network theocommu-

ambiguous

in the activa-

or her own preference. We believe that this property of the task captures a fundamental aspect of the frontal lobe function. If free will can be anchored anywhere in particular in the brain (a proposition we do not necessarily advocate other than after some red wine), is assuredly in the frontal lobes. The neuropsychology lobes should then, of the frontal at least in part, of in the psysuch in-

rists (J. G. Taylor, nication, 1995).

DOUGLAS

based/context-dependent

proces-

F. WAI-F,

sor is more involved in the activation of positive or at least benign affective states? (There is only work of both some interesting on this: Josephs 6. speculative extrapolation5 being

Neuropsychology Quincy Hospital,

References

Department, Quincy, MA

it

Galins hypothesis.6) The right hemispheres ontogenetically cally the more

and phylogenetiprimitive system. and the more

1. Podell K, Lovell M, Zimmerman M, Goldberg E: The Cognitive Bias Task and lateralized frontal lobe functions in males. J Neuropsychiatry Cliii Neurosci 1995; 7:491-501

2. Levy J: Psychobiological
bilateral asymmetry,

implications
in Hemisphere

of
Dis-

become the neuropsychology free will. Yet there is nothing instrumentarium of cognitive chology that would permit

7.

Affective

processing

JOURNAL

OF

NEUROPSYCHIATRY

233