Since the publication of Duval and Wicklund's book A theory of Objective Self Awareness in 1972 research on self-awareness has been extremely fertile. Fifteen years later more than 200 articles had already been published on self-focused attention (Franzoi, 1986). Many dimensions of self-awareness have been examined these past 20 years (for reviews, see Davis & Franzoi, 1991; Franzoi, 1996; Gibbons, 1990; Vallerand & Losier, 1994), among which the cognitive substrates of self-observation -- i.e., what cognitive processes are activated during the acquisition of self-information? In social experimental psychology self-awareness has been defined as the ability to become the object of one's own attention; it refers to a sophisticated form of self-attention which includes self-evaluation, where the individual actively identifies, processes and stores information about the self. This definition thus excludes lower forms of self-awareness -- bodily awareness for instance, and neurological manifestations of its impairment: asomatognosia, neglect, hemisomatognosia, the autotopagnosia denial syndrome, etc. (see Kinsbourne, 1995). Morin (1993; see also Morin & Everett, 1990) proposed that self-awareness is mediated by inner speech: the individual in a state of self-awareness would talk to himself/herself about his/her personal characteristics and behaviors. Some empirical evidence supports this hypothesis: in three independant studies, a significant positive correlation has been found between measures of inner speech and self-awareness (Morin, Everett, Turcotte & Tardif, 1993; Rivest & Khawaja, 1995; Siegrist, 1995). Siegrist (1996) also found that highly self-aware individuals significantly use inner speech more frequently in comparison to less self-aware individuals. This suggests that the more one focuses on the self the more one talks to oneself about oneself, and vice versa.

Many techniques can be used to test the hypothesis of a relation between inner speech and self-awareness (Morin & Everett, 1990). Recent brain research offers a new and exciting avenue, where a comparison between brain areas activated during subvocal speech articulation (inner speech) and during self-awareness tasks can be established in order to determine if regions involved in these two activities correlate. In what follows I will suggest that the brain regions involved in self-awareness are identical to the regions responsible for inner speech. This can be taken as additional evidence of a link between self-talk and self-focus.

The neurological basis of inner speech is fairly well-known. A common assertion is that the neurological underspinnings of inner and public communication are the same (Ramsberger, 1994). The classic work of Luria (1976; Luria & Homskaya, 1963) supports this idea. Evidence from more recent investigations using brain imaging techniques points to the left inferior frontal gyrus. For example, McGuire, Silbersweig, Murray, David, Frackowiak and Frith (1996) recorded the activity of brain areas using PET scan in normal subjects silently articulating sentences. Inner speech was associated with activation of the left inferior frontal area, whereas auditory verbal imagery (the monitoring of one's own speech), induced by asking the subjects to imagine the sentences they generated being spoken to them in another person's voice, was associated with increases in the same region, and in the left premotor cortex, the supplementary motor area and the left temporal cortex (see also McGuire, Silbersweig & Frith, 1996). Similar results have been found in other studies. For instance, McGuire, Silbersweig, Wright, Murray, Frackowiak and Frith (1996) examined brain activation in normal subjects and schizophrenic patients with and without hallucinations. Subjects were asked to silently read single words; all three groups showed increased activity in the left inferior frontal region.

Case studies of patients with speech pathologies lead to less consistent results. In some investigations (e.g., Levine, Calvanio & Popovics, 1982) destruction of the the left inferior frontal area following a stroke has been associated with a complete loss of inner speech. However, in other studies different regions seem to be involved as well. For instance, we know that anarthria (a total inability to articulate speech caused by cortical, subcortical and/or brain stem lesions) negatively affects inner speech (Cubelli & Nichelli, 1992). Also, inner speech, defined by Verstichel, Bourak, Font and Crochet (1997) as the capacity to to mentally generate phonological representation of words, is interrupted by lesions to Broca's area -- an observation that supports the view that spoken and inner speech are mediated by similar brain regions. This apparent lack of consistency is likely due to the use of different measures of inner speech, and to the fact that in many cases the localization of the insult is tentative at best.

This last observation also applies to most studies that have been conducted in order to correlate brain activity with self-awareness. Many investigations indicate that frontal lobes lesions are associated with self-awareness deficits (e.g., Jurado, Junque, Vendrell, Treserras & Grafman, 1998; Miller, 1991; Stuss, 1991a, b). Disturbances of impaired self-awareness are also common after traumatic brain injury, which very frequently involves frontal lobes damage (Prigatano, 1991; Prigatano, Ogano & Amakusa, 1997). Duffy (1995) states -- without really explaining why -- that the dorsal convexity and orbitofrontal areas of the frontal lobe are responsible for monitoring the individual's intrapersonal world. In an examination of patients with CT/MRI confirmed focal frontal lesions, Leduc, Herron, Greenberg, Eslinger and Grattan (1999) found deficits in social self-awareness (measured with a validated scale) in patients with orbital frontal damage.

These observations, while pointing into the right direction, still provide information which is too vague to support my contention that inner speech and self-awareness are mediated by identical brain areas. The most pertinent study [1] in that respect has been conducted by Craik, Moroz, Moscovitch, Stuss, Winocur, Tulving and Kapur (1999). This team measured brain activity in normal subjects working on a self-referential encoding task. Participants were requested to judge how well trait adjectives described them by pressing response keys while relative regional cerebral blood flow was being measured. This task clearly represents a self-awareness task since it involves thinking about oneself. Control tasks consisted in three non self-referential exercices -- judging how well trait adjectives described a public figure, how socially desirable the trait adjectives were, and how many syllables there were in each adjective. Results show that the self-referential encoding task produced an increased activity in the left medial aspect of the superior frontal gyrus and in the left inferior frontal gyrus.

I propose that taken together, these observations provide additional support to the notion that the individual in a state of self-awareness actually uses inner speech. Since inner speech and self-awareness seem to share a common neurological basis (the left inferior frontal region), it is highly tempting to conclude that these two operations are deeply linked. The ultimate test of this hypothesis would be to measure brain activity in subjects confronted to self-focusing stimuli -- stimuli like a mirror, that have been repeatedly found to produce a state of self-awareness (Carver & Scheier, 1978).

NOTE [1]. In a recent article Gallup (1999) claims that self-awareness may correlate with activity in the right prefrontal cortex. He cites a study conducted by Keenan, Mccutcheon, Freund, Gallup, Sanders & Pascual-Leone (1999), in which brain activity was recorded in subjects making self-recognition tasks -- subjects viewing their own faces displayed significant changes in electrical potentials in the right prefrontal cortex. While these results are informative I believe that they do not support Fallup's view. Self-recognition (a manifestation of self-awareness) may indeed involve the right prefrontal cortex, but this does not mean that this area represents the neurological substrate of the cause of self-recognition -- self-awareness.

Verstichel, Patrick, Bourak, C.; Font, V., & Crochet, G. (1997). Langage intérieur après lésion cérébrale gauche: Etude de la représentation phonologique des mots chez des patients aphasiques et non aphasiques [Inner speech and left brain damage: Study of the phonological analysis of words in aphasic and non-aphasic patients]. Revue de Neuropsychologie, 7(3), 281-311.