Higher-Level Cognition - Computational Explorations in Cognitive Neuroscience - page 397

Fluency Executive Control

Another important angle on the issue of flexibility
comes from the study of fluency tasks with frontal pa-
tients. These tasks typically require the subject to gen-
erate as many different responses to a given stimulus
as possible, for example by naming as many different
animals that they can think of in a limited amount of
time, or by generating as many different words that be-
gin with a given letter. Interestingly, frontal patients do
not show deficits on the animal-naming version of the
fluency task (Newcombe, 1969), but do show deficits
on the word initial-letter version (Milner, 1964). These
findings show that the deficit is not necessarily asso-
ciated with the control aspects of the generation task,
but rather with the arbitrary nature of the initial-letter
version. Thus, when the responses are all very famil-
iar and fit within a well-established category (e.g., ani-
mals), posterior areas appear to be capable of generating
the responses. In contrast, one is not often required to
generate words based on initial letters, so this requires
akindofflexibility in adapting to the novel demands of
the task.
Another fluency task, the Alternative Uses test, ap-
pears to place even greater demands on the frontal cor-
tex because it is more sensitive to frontal damage (But-
ler, Rorsman, Hill, & Tuma, 1993). This test requires
the subject to generate as many atypical uses for a com-
mon object as possible (e.g., using newspaper to make a
hat, wrap fish, start a fire). In addition to requiring novel
responses, these responses must overcome the prepotent
responses of the typical uses for an object.
The continuum of frontal involvement in these flu-
ency tasks is consistent with the activation-based model
of frontal function. In short, top-down support from the
frontal cortex is needed to support the unfamiliar pro-
cessing involved in the initial-letter word fluency task,
but not in the familiar animal-naming task. The Alterna-
tive Uses test is even more demanding of frontal func-
tion because strong prepotent responses must be over-
come, presumably via strong top-down activation from
the frontal cortex.

Deficits in executive control associated with frontal le-
sions are commonly observed, but can be difficult to
characterize due to the complex nature of these pro-
cesses. One important aspect of executive control is
goal-directed behavior. As early as Bianchi (1922),
frontal lesions have been characterized as impairing the
ability to organize a series of actions around a com-
mon goal. For example, Bianchi described a frontally
lesioned monkey that grabbed a door handle but then
failed to open the door because it became engaged in
looking at the handle; then it just sat on the handle. It
appears that actions under frontal damage are more of-
ten reactions to environmental cues rather than deliber-
ate and goal-directed. Lhermitte (1986) has described
an environmental dependency syndrome accompanying
frontal damage. For example, a patient visiting a physi-
cian's home reacted to seeing a hammer, nail, and pic-
ture by hanging the picture on the wall, and another saw
a set of dirty dishes and immediately began washing
them!
In addition to not seeming to be able to generate goal-
driven behavior themselves, frontal patients have diffi-
culty following explicit task instructions, even for rel-
atively simple everyday tasks. For example, Shallice
and Burgess (1991) described frontal trauma patients
who were given explicit instructions for a sequence of
everyday tasks (shopping for a small set of items, keep-
ing an appointment, and collecting information). These
patients, who scored above average on an IQ test, nev-
ertheless failed to remain focused on the tasks, and be-
came distracted by various intervening events. A sim-
ple lack of memory for the goals can be ruled out as an
explanation for their failure because they were given a
written list of the tasks.
Laboratory problem solving tasks can provide some-
what more controlled tests of frontal executive func-
tions. One such task that has been studied with frontal
patients is the Tower of London task (Shallice, 1982).
In this task, a variant of the Tower of Hanoi, colored
balls placed on a set of three sticks must be moved one
at a time to achieve a specified goal state where the balls
are in a specific arrangement on the sticks. Because
the sticks are of varying heights so that they can only