There is considerable variability among epilepsy centers in the method
s and interpretations of the intracarotid amobarbital procedure. Promi
nent among these differences is the determination of language represen
tation and assessment of language functions, Some centers rely on spee
ch arrest following amobarbital injection as a marker for language rep
resentation, whereas other centers examine verbal output for the prese
nce of aphasic errors. The present study assessed the pattern of langu
age recovery following amobarbital injection in epilepsy patients who
were candidates for temporal lobectomy. Language recovery from dominan
t hemisphere injection (left or right) followed a stereotypical progre
ssion, with 71.8% of patients showing return of vocalization followed
by return of naming and comprehension. Repetition deficits with paraph
asic errors persisted the longest (mean = 12'30 ''), with a conduction
aphasia persisting after the acute global aphasia resolved. Although
two patients interpreted as left hemisphere language dominant were mut
e following right hemisphere injection, all language functions were in
tact immediately upon resumption of vocalization and they showed no ot
her signs of aphasia such as paraphasias or anomia. Possible explanati
ons for serial language recovery and persistent conduction aphasia are
discussed. These findings have significant implications for the deter
mination of cerebral language dominance. (C) 1997 Academic Press.