Prosody consists of alterations in pitch, stress, and duration across words, phrases, and sentences. These same parameters are defined acoustically as fundamental frequency, intensity, and timing. It is the variation in these parameters that not only provides the melodic contour of speech, but also invests spoken language with linguistic and emotional meaning. Prosody is thus crucial to conveying and understanding communicative intent.

The term “aprosodia” was first used by Monrad-Krohn (1947) to describe loss of the prosodic features of speech. It resurfaced in the 1980s in the work of Ross and his colleagues to refer to the attenuated use of and decreased sensitivity to prosodic cues by right hemisphere damaged patients (Ross and Mesulam, 1979; Ross, 1981; Gorelick and Ross, 1987).

Prosodic deficits in expression or comprehension can accompany a variety of cognitive, linguistic, and psychiatric conditions, including dysarthria and other motor speech disorders, aphasia, chronic alcoholism, schizophrenia, depression, and mania, as well as right hemisphere damage (RHD) (Duffy, 1995; Myers, 1998; Monnot, Nixon, Lovallo, and Ross, 2001). The term aprosodia, however, typically refers to the prosodic impairments that can accompany RHD from stroke, head injury, or progressive neurologic disease with a right hemisphere focus. Even the disturbed prosody of other illnesses, such as schizophrenia, may be the result of alterations in right frontal and extrapyramidal areas, areas considered important to prosodic impairment subsequent to RHD (Sweet, Primeau, Fichtner et al., 1998; Ross et al., 2001).

The clinical presentation of expressive aprosodia is a flattened, monotonic, somewhat robotic, stilted prosodic production characterized by reduced variation in prosodic features and somewhat uniform intersyllable pause time. The condition often, but not always, accompanies flat affect, a more general form of reduced environmental responsivity, reduced sensitivity to the paralinguistic features of communication (gesture, body language, facial expression), and attenuated animation in facial expression subsequent to RHD. Aprosodia can occur in the absence of dysarthria and other motor speech disorders, in the absence of depression or other psychiatric disturbances, and in the absence of motor programming deficits typically associated with apraxia of speech. Because it is associated with damage to the right side of the brain, it usually occurs in the absence of linguistic impairments (Duffy, 1995; Myers, 1998).

Expressive aprosodia is easily recognized in patients with flat affect. Deficits in prosodic perception and comprehension are less apparent in clinical presentation. It is important to note that receptive and expressive prosodic processing can be differentially affected in aprosodia.

First observed in the emotional domain, aprosodia has also been found to occur in the linguistic domain. Thus, patients may have problems both encoding and decoding the tone of spoken messages and the intention behind the message as conveyed through both linguistic and emotional prosody.

In the acute stage, patients with aprosodia are usually unaware of the problem until it is pointed out to them. Even then, they may deny it, particularly if they suffer from other forms of denial of deficit. Severity of neglect, for example, has been found to correlate with prosodic deficits (Starkstein, Federoff, Price et al., 1994). In rare cases, aprosodia may last for months and even years when other signs of RHD have abated. Patients with persistent aprosodia may be aware of the problem but feel incapable of correcting it.

Treatment of aprosodia is often limited to training patients to adopt compensatory techniques. Patients may be taught to attend more carefully to other forms of emotional expression (e.g., gesture, facial expression) and to signal mood by explicitly stating their mood to the listener. There has, however, been at least one report of successful symptomatic treatment using pitch biofeedback and modeling (Stringer, 1996). Treatment has been somewhat limited by uncertainty about the underlying mechanisms of aprosodia. It is not clear the extent to which expressive aprosodia is a motor problem, a pragmatic problem, a resource allocation problem, or some combination of conditions. Similarly, it is not clear whether receptive aprosodia is due to perceptual interference in decoding prosodic features, to restricted attention (which may reduce sensitivity to prosodic cues), or to some as yet unspecified mechanism.

Much of the research in prosodic processing has been conducted to answer questions about the laterality of brain function. Subjects with unilateral left or right brain damage have been asked to produce linguistic and emotional prosody in spontaneous speech, in imitation, and in reading tasks at the single word, phrase, and sentence level. In receptive tasks they have been asked to determine the emotional valence of expressive speech and to discriminate between various linguistic forms and emotional content in normal and in filtered-speech paradigms. Linguistic tasks include discriminating between nouns and noun phrases based on contrastive stress patterns (e.g., greenhouse versus green house); using stress patterns to identify sentence meaning (Joe gave Ella flowers versus Joe gave Ella flowers); and identifying sentence types based on prosodic contour (e.g., the rising intonation pattern for interrogatives versus the flatter pattern for declaratives).

The emphasis on laterality of function has helped to establish that both hemispheres as well as some subcortical structures contribute to normal prosodic processing. The extensive literature supporting a particular role for the right hemisphere in processing content generated the central hypotheses guiding prosodic laterality research. The first hypothesis suggests that affective or emotional prosody is in the domain of the right hemisphere (Heilman, Scholes, and Watson, 1975; Borod, Koff, Lorch et al., 1985; Blonder, Bowers, and Heilman, 1991). Another hypothesis holds that prosodic cues themselves are lateralized, independent of their function (emotional or linguistic) (Van Lancker and Sidtis, 1992). Finally, lesion localization studies have found that certain subcortical structures, the basal ganglia in particular, play a role in prosodic processing (Cancelliere and Kertesz, 1990; Bradvik, Dravins, Holtas et al., 1991). Cancelliere and Kertesz (1990) speculated that the basal ganglia may be important not only because of their role in motor control, but also because of their limbic and frontal connections which may influence the expression of emotion in motor action.

Research findings have varied as a function of task type, subject selection criteria, and methods of data analysis. Subjects across and within studies may vary in terms of time post-onset, the presence or absence of neglect and dysarthria, intrahemispheric site of lesion, and severity of attentional and other cognitive deficits. With the exception of site of lesion, these variables have rarely been taken into account in research design. Data analysis has varied across studies. In some studies it has been based on perceptual judgments by one or more listeners, which adds a subjective component. In others, data are submitted to acoustic analysis, which affords increased objective control but in some cases may not match listener perception of severity of impairment (Ryalls, Joanette, and Feldman, 1987).

In general, acoustic analyses of prosodic productions by RHD patients supports the theory that prosody is lateralized according to individual prosodic cues rather than according to the function prosody serves (emotional versus linguistic). In particular, pitch cues are considered to be in the domain of the right hemisphere. Duration and timing cues are considered to be in the domain of the left hemisphere (Robin, Tranel, and Damasio, 1990; Van Lancker and Sidtis, 1992; Baum and Pell, 1997).

Research suggests that reduced pitch variation and a somewhat restricted pitch range appear to be significant factors in the impaired prosodic production of RHD subjects (Colsher, Cooper, and Graff-Radford, 1987; Behrens, 1989; Baum and Pell, 1997; Pell, 1999a). RHD patients are minimally if at all impaired in the production of emphatic stress. However, they may have an abnormally flat pitch pattern in declarative sentences, less than normal variation in pitch for interrogative sentences, and may produce emotionally toned sentences with less than normal acoustic variation (Behrens, 1988; Emmory, 1987; Pell, 1999). Pitch variation is crucial to signaling emotions, which may explain why impaired production of emotional prosody appears particularly prominent in aprosodia. Interestingly, in the case of tonal languages (e.g., Chinese, Thai, and Norwegian) in which pitch patterns in individual words serve a semantic role, pitch has been found to be a left hemisphere function (Packard, 1986; Ryalls and Reinvang, 1986; Gandour et al., 1992).

Prosodic perception or comprehension deficits associated with aprosodia tend to follow the pattern found in production. Non-temporal properties such as pitch appear to be more problematic than time-dependent properties such as duration and timing (Divenyi and Robinson, 1989; Robin et al., 1990; Van Lancker and Sidtis, 1992). For example, Van Lancker and Sidtis (1992) found that right-and left-hemisphere-damaged patients used different cues to identify emotional stimuli. Patients with RHD tended to base their decisions on durational cues rather than on fundamental frequency variability while left-hemisphere-damaged patients did the opposite. These data suggest a perceptual, rather than a functional (linguistic versus emotional), impairment. Although a study by Pell and Baum (1997) failed to replicate these results, the data are supported by data from dichotic listening and other studies that have investigated temporal versus time-independent cues such as pitch information (Chobor and Brown, 1987; Sidtis and Volpe, 1988; Divenyi and Robinson, 1989; Robin et al., 1990).

Almost all studies of prosodic deficits have focused on whether unilateral brain damage produces prosodic deficits, rather than describing the characteristics of prosodic problems in patients known to have prosodic deficits. The body of laterality research has established that prosodic deficits can occur in both left as well as right hemisphere damage, and has furthered our understanding of the mechanisms and differences in prosodic processing across the hemispheres. However, the focus on laterality has had some drawbacks for understanding aprosodia per se. The main problem is that while subjects in laterality studies are selected for unilateral brain damage, they are not screened for prosodic impairment. Thus, the data pool on which we rely for conclusions about the nature of RHD prosodic deficits consists largely of subjects with and subjects without prosodic impairment. The characteristics of aprosodia, its mechanisms, duration, frequency of occurrence in the general RHD population, and the presence/absence of other RHD deficits that may accompany it have yet to be clearly delineated. These issues will remain unclear until a working definition of aprosodia is established and descriptive studies using that definition as a means of screening patients are undertaken.

See also prosodic deficits, right hemisphere language and communication functions in adults; right hemisphere language disorder.