Stanford W. Gregory

Voice pitch and amplitude convergence as a metric of quality in dyadic interviews

When a man speaks to an assembly, in Durkheim’s words (1961, p. 241), ‘the demon of oratorical inspiration’ oversees his speech, ‘His language has a grandiloquence that would be ridiculous in ordinary circumstances; his gestures show a certain domination; his very thought is impatient of all rules, and easily falls into all sorts of excesses . . . Now this exceptional increase of force is something very real; it comes to him from the very group which he addresses . . . It is no longer a simple individual who speaks; it is a group incarnate and personified.’ Though Durkheim’s comments reflect a larger community interaction and a chiliastic context differing from ordinary circumstances, a similar adaptative control is imposed on individuals in more ordinary micro-interactive circumstances. The public address with its theatricality and associated exaggerations illustrates well, however, the ways in which persons mutu~ly adapt. ’ The ‘common conscience’ or ‘social similitude’ (Durkheim, 1964, p. 80) found in Durkheim’s notion of mechanical solidarity and exemplified in his ‘demon of oratorical inspiration’ provides the micro-organization leading to a ‘determinate system which has its own life . . . independent of the particular conditions in which individuals are placed’ (Durkheim, 1964, pp. 79-80), a statement epitomizing his social psychological conception of interaction.

Analysis of fundamental frequency reveals covariation in interview partners' speech

Spectral analysis of the fundamental frequency band in speech of interview partners reveals covariance of voice energy levels and thus a possible form of rudimentary social synchrony. Though this phenomenon of paralinguistic covariance has been observed by a number of researchers, techniques for examining it have not been exploited using modern instruments for Fast Fourier Transform analysis nor has the effect been statistically explicated in depth. This study reports on research showing that the acoustic signal conveying covariance information resides in the fundamental frequency band of the speech spectrum, a band which in itself has traditionally been considered inconsequential to perceived intelligible verbal communication. Acoustic signals from speech of four interview partners interacting with the author were analyzed in three frequency settings—2 KHz., .5 KHz., and .2 KHz.—to target the low frequency band in which the covariance phenomenon is most coherently transmitted. Analysis was performed using a dedicated instrument, the Fast Fourier Transform analyzer, to obtain spectra, and spectral output was processed statistically to produce relationships showing the covariance effect. Also, this paper addresses means by which potency of influences from technical artifacts, as introduced by instrumentation, interview environment, or statistical protocols are appraised. Research conclusions present an efficient, lucid, and reliable method for analyzing the paralinguistic mode of nonverbal behavior, and, in addition, offer evidence that the nonverbal, vocal channel of communication carries a signal embodying a semantic message. These conclusions raise some crucial questions regarding the influence of a hitherto unreported function of the fundamental frequency in voices of interview partners.

Spectral Analysis of Candidates' Nonverbal Vocal Communication: Predicting U.S. Presidential Election Outcomes*

Fast Fourier Transform acoustic analysis of the fundamental frequency of candidates’ voices in 19 nationally televised U.S. presidential debates from the eight elections including debates held since 1960, in conjunction with subsequent factor analysis, shows that this nonverbal frequency, below .5 kHz, can reveal the debating candidates’ relative social dominance. Further analysis presents evidence that the candidates’ nonverbal vocalizations offer a precise metric of their relative dominance or commanding presence in the presidential campaign: when this metric is compared statistically with the candidates’ popular vote percentages for the U.S. presidency, it accurately predicts the popular vote outcomes in all of those eight elections.

Sounds of Power and Deference: Acoustic Analysis of Macro Social Constraints on Micro Interaction

In Goffmans introduction to Interaction Ritual, he implores sociologists “to identify the countless patterns and natural sequences of behavior occurring whenever persons come into one anothers immediate presence.” This article, in addressing Goffmans entreaty, amalgamates a number of sociological theories on interaction ritual from Durkheim and Goffman with others in contemporary literatures in the sociology of emotions and social psychology to describe a set of “patterns and natural sequences of behavior” imbedded in voices of interaction partners. These patterns are shown to establish an element of social structure, and this research describes an objective method of identifying and measuring this element that previously was known on a qualitative basis alone. A social status ordering pattern founded upon interpersonal deference and power relations is identified through analysis of acoustic energy levels in voices of interview partners. Nonverbal measures are performed using a dual-channel Fast Fourier Transform analyzer and results are explained using a carrier spectrum model. This report sets forth an interesting new way of objectively measuring the macro social constraints on micro interaction.

Examining Medical Interview Asymmetry Using the Expectation States Approach

In this study we examine medical interview asymmetry using the expectation states approach. Physicians lead clinical interviews because of a feature inherent in those interviews, namely the status difference between doctor and patient. This power differential varies: it is greatest when the biomedical aspects of the interview are emphasized. These observations are consistent with status characteristics theory (SCT), which is based on the expectation states approach to understanding the emergence of power-prestige orders in groups facing shared tasks. From an SCT perspective, when the required scope conditions are met the status characteristics of doctor and patient trigger expectation states that result in inequalities relevant to the biomedical tasks of the interview. We examine interactions between medical students and standardized patients from the perspective of SCT. We observe the emergence of vocal spectrum inequalities when the interview task is biomedical. Other nonverbal behavioral outcomes emerge as well, which are consistent with the asymmetry literature.

Social psychological implications of voice frequency correlations: analyzing conversation partner adaptation by computer

Several earlier papers have discussed the results of a technique for analyzing paralinguistic adaptation of partners in conversation. The previous research showed that in a group of eleven tape-recorded dyadic conversations, where the actual conversation partners have been separated and then scrambled together, a set of computer routines is capable of rematching a digitized version of the actual conversation partners through an analysis of voice frequency levels. In another paper based upon the same data set and making use of the same techniques of analysis, it was found that the adaptation phenomenon is also capable of quantitatively and objectively selecting out from a scrambled sample of group and nongroup members, specific conversations associated with persons who share membership in a consolidated group. Members of a consolidated group in comparison with desparate individuals engage in a kind of paralinguistic crypto-communication constituting a code signifying their membership. These sociolinguistic techniques present innovative new ways of researching a multitude of general social psychological areas such as conversational affectivity, conflict and effectiveness, which obviously specifically influence social interaction between patient and therapist, non-native speakers and native speakers, salespeople and customers, professional interviewers and subjects, as well as a multitude of other diverse applications. Though ideas dealt with here are not difficult to understand, some of the techniques used are less accessible. This paper explicitly sets forth methods used in this type of research in a less technical manner, in an effort to make techniques more accessible to the social psychological researcher.

Comprehending the Neurological Substratum of Paraverbal Communications: The Invention of SplitSpec Technology

An earlier book chapter (Gregory 1999) reviewed the specific background of our investigations into the social significance and neurology of the lower speech frequency in human communication. This lower speech frequency, termed the paraverbal frequency (beneath.5 kHz), was found to convey important nonverbal social information, and the previously published chapter outlined the methodology and results of numerous experiments showing how the paraverbal signal differs from the verbal; specifically, how it acts as an elemental mechanism of social status accommodation and social convergence between conversation partners. The present chapter will continue from where the last chapter left off and aims first to merge our past results and observations with a theoretical account making use of pertinent findings from physical anthropology, cognitive psychology, and neurology. With the theoretical background established, we will then review how our research led to the development of a technological innovation called “SplitSpec Technology,” and how this innovation will fit into the future of human electronic communications.

Nonverbal Vocal Adaptation and Audience Perceptions of Dominance and Prestige

Fast Fourier transform (FFT) analysis of speech is a useful tool for measuring the social organization of micro interactions. Past research suggests that “adaptation” in the lower nonverbal vocal frequencies may be associated with bystander perceptions of dominance and prestige, but these factors have not been compared in a single study. Furthermore, it is unclear whether nonverbal vocal adaptation has an independent effect on bystander perceptions of dominance or prestige in naturalistic settings. To explore these questions, we aired manipulated versions of interviews from Piers Morgan Live to panels of undergraduates who then rated the dominance and prestige of the host and his guest. We find that nonverbal vocal adaptation is uniquely associated with dominance perceptions. We also find that nonverbal vocal adaptation does not have an effect on dominance perceptions that is independent of other cues. A step-by-step guide to help others use FFT voice analysis is provided.

Social Psychology, Past and Present-An Integrative Orientation.

Contents: Early Thinking About Social Persons. Influences from the Recent Past. The Formative Years of Scientific Social Psychology. Contemporary Social Psychology: The Early Decades. The Crisis in Social Psychology. Current Trends in Social Psychology. An Integrative Theoretical Orientation.