• No se han encontrado resultados

Within the cortex of the superior temporal sulcus (STS), populations of cells have been studied that respond more to the sight of faces than to a variety of other simple and complex stimuli (eg. Bmce et al 1981, Desimone et al 1984, Gross et al 1972, Perrett et al 1982,1984,1985). Some evidence does exist that special structures of the brain control the execution and understanding of facial identification. Microelectrode recordings perfonned in awake monkeys indicate that face specific neuronal networks exist in the primate brain located in cortical areas bordering on the central region of the superior temporal sulcus of monkey brains (Perrett et al 1982, 1984; RoUs 1984; Desimone et al

These cells have also been shown to be tolerant to changes in viewing conditions such as size, orientation and position. Some cells have also been found which are sensitive to the identity of faces (eg. Baylis et al 1985, Perrett et al 1984,1986,1987, Rolls 1984). In this case, cells respond more to the face of one individual (familiar to the monkey) than to other equally familiar faces. These cells also generalise across viewing conditions such as orientation, viewing distance and expression. Neuroanatomical studies have shown that these face responsive cells occur in particularly high concentration within two regions of the STS, areas TPO and PGa (eg. Perrett et al 1986,1987,1988). They are also found to a lesser degree in other parts of the inferior temporal cortex. The face responsive cells are found in patches extending vertically through the thickness of the cortex and in a tangential direction across the cortical surface. It should be noted that the STS does not only contain cells responsive to faces. Other cells within the region respond to

somatosensory infonnation or different classes of movement. Thus the area is not exclusive for face processing, but the evidence suggests a face processing subsystem within one brain area. Cells sensitive to faces have also been reported in other brain areas (eg. frontal cortex, parietal cortex amygdala and brain stem) but the highest concentration of these cells is found in the temporal cortex (see Fig. 3.16 from Perrett et al 1991). Another source of investigation which has indicated face specific brain mechanisms are humans who exhibit neuropsychological deficits. This research will be briefly reviewed. 3.5 NEUROPSYCHOLOGICAL DEFICITS

Deficits in face processing were first formally described by Bodamer (1947) who used the term 'prosopagnosia' to describe the inability to recognise familiar faces. Since then, well documented impairments in the recognition of familiar faces have been found without disturbances in the linguistic domain (eg. Gmsser 1984, Jeeves 1984, Young et al

%

g

8

S

1

j

!

♦\2/

\ ^ A Î

; î 'i 1 7 1 1

4to

j

-V

<3

l

(£ >- X OC 1^;

# 4 *

- O

îs'ït!

1988). There have been cases reported of prosopagnosics who can perfonn nonnally in the matching of unfamiliar faces (Landis et al 1979, De Haan et al 1987). There is also evidence that unfamiliar face recognition impairment can occur with no specific deficit in familiar face recognition (eg. Warrington and James 1967, De Renzi et al 1968).

Bauer (1984) and Tranel (1985) have shown that prosopagnosics who are unable to discriminate between familiar and unfamiliar faces in a behavioural task do produce galvanic skin responses which show a clear discrimination between familiar and

unfamiliar faces. This covert recognition of familiar faces has been proposed as evidence for two separate pathways for processing facial information.

De Haaii et al (1987) also report a case of covert recognition. The patient (PH) was unaware that any stimulus face were familiar to him, but he responded faster when task decisions involved familiar rather than unfamiliar faces.

Cases of prosopagnosia have rarely been found without the occurrence of other deficits (Meadows 1974, Whiteley and Warrington 1977), and it is common to find that

prosopagnosics also show deficits in recognition of other classes of objects. Damasio et al (1982) suggest this is evidence that prosopagnosia is not a face specific deficit but an inability to distinguish between exemplars of a certain category. Cases have been found, however, where prosopagnosics show little evidence of object agnosia (Whiteley and Warrington 1977, Newcombe 1989, De Renzi 1986).

Cases of prosopagnosia give some support for the notion of stimulus selective processmg mechanisms but as noted this evidence is controversial because most patients have a variety of recognition impamnents. Damage in prosopagnosia is rarely localised to one

functional system, so this does not preclude the possibility that there are specific systems for face processing.

The issue of whether there exist evoked potentials which are selectively responsive to facial stimuli is explored later in this thesis, the first experiments to be reported are concerned with how repeating a stimulus influences waveforms evoked in monkey brains.

CHAPTER FOUR

EXPERIMENT ONE: E.R.Ps RECORDED FROM MONKEYS USING FAMILIAR PICTURES IN A MATCH-TO-S AMPLE TASK

CHAPTER FOUR

EXPERIMENT ONE: E.R.Ps RECORDED FROM MONKEYS USING FAMILIAR PICTURES IN A MATCH-TO-SAMPLE TASK

J 4.1 INTRODUCTION

In the following chapter, an investigation of evoked potential responses recorded from monkeys is described. The main question of interest is can the E.R.P. repetition effect previously described in Chapter 2 be replicated in monkeys using non-verbal material? In human studies, E.R.Ps to repeated words show a sustained positive shift in comparison with non-repeated words (Rugg 1985,1987). This modulation of evoked potentials when items are repeated has led to many insights into memory organisation (See Chapter 2 for review). The following investigation extends the enquiry to monkey evoked potentials. Most E.R.P. studies of repetition effects in humans have employed words as stimulus items, as they can be manipulated along many dimensions - familiarity, frequency, orthographically legal/illegal etc. In the study reported here using monkeys as subjects, it was necessary to find a comparable set of stimulus items. The stimuli chosen to this end were pictures. The task element of the study paralleled the human studies reported by

Rugg and his collaborators (1985,1987, Rugg and Barrett 1988) as far as was possible. î j Rugg (1987) adopted a paradigm in which E.R.P. repetition effects could be studied with ! items to which no overt response was required but which were subject to an implicit

lexical decision. Similarly, in this study, no overt response was required to the critical items but the monkeys made match/non-match decision to every stimulus. It was hoped by using this paradigm to extend understanding of E.R.P. repetition effects across species.

Some of the experiments reported here which involve the recording of evoked potentials from monkey brains are innovative and are presented as pilot studies. Throughout the

experiments conducted, there were many difficulties encountered due to the length of training time, inconsistent performance by the monkeys and electrical problems. The number of recording sites was restricted in some experiments due to faulty electrodes which could not be repaired non-invasively. This renders the interpretation of the data problematic from the point of view of source locations of potentials. The experiments are

also based on recordings from two experimental subjects, and use of one of these subjects was limited by medical problems. Despite these difficulties, some interesting and

significant results were obtained which point the way to future investigations.

4.2 METHODS