Social neuroscience is more than the study of the human brain: The legacy of John Cacioppo

ABSTRACT

John Cacioppo passed away in 2018, leaving a legacy of profound methodological, theoretical, and inferential contributions to social neuroscience. This paper serves as an introduction to the nine articles that comprise this special issue in honor of John Cacioppo’s work in social neuroscience. Although he made many contributions to psychology, here we briefly review four milestones in Cacioppo’s career that had important implications specifically for the development of social neuroscience today: (1) an early research focus on cardiovascular and facial EMG measurement, (2) the training of others, (3) the importance of sound inference, and (4) the definition of social neuroscience. In sum, we argue that John Cacioppo envisioned social neuroscience as having multiple levels of explanation and requiring multiple kinds of physiological evidence. It is not all just the brain!

When John T. Cacioppo passed away on 5 March 2018 at the age of 66, he left an enormous legacy to the field of social neuroscience, which included being one of the founders of this interdisciplinary field, as well as making several important contributions to the theory, methodology, and range of topics that are covered in it. The articles in this special issue reflect the breadth and scope of John’s vision of social neuroscience – that it should include human and non-human research, a wide range of methods, and, of course, multiple levels of analysis. The contributors to this issue have all been influenced directly or indirectly by this vision. Other writers have provided fuller accounts of his remarkable life, so we do not cover his biography here, but highlight some key aspects of his work that are most relevant to this special issue. Each of us were greatly influenced by John. Eric Vanman was an undergraduate in John’s laboratory at the University of Iowa. Arvid Kappas was a life-long colleague who especially appreciated John’s advice early in his career. Tiffany Ito was a post-doc in John’s lab at the Ohio State University. Here we highlight four milestones from John’s career that have informed the development of social neuroscience.

Early focus on cardiovascular activity and facial electromyography

John was trained as a social psychologist at Ohio State in the 1970s. One of the strengths of the graduate program at that time was its research on attitudes and persuasion. John, along with his classmate, Richard Petty, thus developed strong interests in this area. They soon began collaborating on several key studies of attitude change. John’s doctoral dissertation was published in Journal of Personality and Social Psychology (Cacioppo, 1979), and that research consists of his first foray into investigating a social psychological phenomenon with psychophysiology. The title of the article, “Effects of exogenous changes in heart rate on facilitation of thought and resistance to persuasion,” reveals the novelty of his approach. Working with a cardiologist, John recruited a group of patients who had implanted cardiac pacemakers. In two experiments, these participants read persuasive messages while their pacemakers were manipulated to increase or maintain their basal heart rate. John found that, even though participants could not detect changes in their cardiovascular activity, heart rate acceleration led to more counterarguments and greater resistance to attitude change compared to baseline, which supported hypotheses he had derived from early work by Lacey (1967). Social psychologists at the time had been concerned with the perception of autonomic nervous system (ANS) activity as a moderator of social cognition, but John’s study was one of the first to examine actual, unperceived changes in the ANS on cognitive responses.

Similarly, John’s interests in the psychophysiology of attitudes and persuasion led him to another important methodology that he and his collaborators frequently employed in the following decade–facial electromyography (EMG). Although today facial EMG is typically used to make inferences about underlying affective states, John’s initial interests were no doubt inspired by earlier research on the measurement of perioral activity as a marker of cognitive processes. His first paper using this method (Cacioppo & Petty, 1979) cited Sokolov’s (1969) work that demonstrated “bursts of covert oral EMG” prior to learning a complex matrix, and McGuigan’s (1970) research showing that oral EMG was greater for non-proficient readers than proficient readers. In their study, Cacioppo and Petty presented participants a pro-attitudinal, counter-attitudinal, or neutral message while oral EMG was recorded. Whenever forewarned that they were to receive a counter-attitudinal message, participants exhibited more oral EMG activity than when they did not receive the forewarning, which Cacioppo and Petty interpreted as consistent with participants’ greater information processing as they encountered a discrepant message.

Throughout the 1980s, John and his colleagues continued to use facial EMG in studies of information processing and persuasion (e.g., Cacioppo et al., 1988; Cacioppo & Petty, 1981; Cacioppo et al., 1986, 1985; Petty & Cacioppo, 1983). Their use of facial EMG to measure cognitive processes, however, gradually gave way to its use as a measure of affective processes that could be continuously monitored throughout the testing situation while the recording apparatus remained relatively unobtrusive. This later approach was not novel in itself, as facial EMG was already being used to investigate emotion by other researchers (e.g., Dimberg, 1982; Fridlund et al., 1984; Schwartz et al., 1976; Vaughan & Lanzetta, 1981). However, few researchers applied the measure as broadly and systematically as John and his collaborators, who used facial EMG as a tool to investigate social, emotional, and clinical questions. One important example of this research is Cacioppo et al. (1986), in which the researchers demonstrated that zygomaticus major (cheek) and corrugator supercilii (brow) EMG activity increased to positive and negative stimuli, respectively. Such changes occurred at a covert level. That is, an analysis of overt facial activity failed to reveal any observable changes in expression. A second example is Cacioppo et al. (1988), in which participants talked about their lives to an interviewer while facial EMG from the brow region was recorded. The video of the interview was then played back to the participant. The tape was stopped at predetermined points and the participant was prompted to report anything else that was going on in their mind at that moment. Cacioppo et al. found that inconspicuous brow EMG responses covaried with subtle variations in emotions during the interview, even though observers trained in facial measurement and clinical judgment could not detect such changes overtly. Findings like these from the Cacioppo lab paved the way for others to use facial EMG to measure affective responses that a participant might be unable or unwilling to report.

Leading and training others

During the 1980s John became a leader in what was then called social psychophysiology and, that later, in the 1990s, developed into the field of social neuroscience. His leadership took two forms – as the editor of seminal handbooks that introduced and standardized the field, and as a teacher leading intensive four-week summer training workshops for established researchers in social psychology. In the first case, as a handbook editor, John co-edited (with Petty) Social Psychophysiology: A Sourcebook (Cacioppo & Petty, 1983), which included chapters on methodology as well content (e.g., interpersonal processes, health, attitudes, emotions) by leaders in that nascent field. Later, with Lou Tassinary, John co-edited the mammoth Principles of Psychophysiology: Physical, Social and Inferential Elements (Tassinary & Cacioppo, 1990), which is now in its fourth edition and titled the Handbook of Psychophysiology(and later co-edited with Gary Berntson). The chapters in these books are gold standards in methodology and inference, and no doubt became required reading for many graduate students in subsequent decades.

John’s role in training social psychophysiologists is perhaps less well known, but in 1986 he received funding from the National Science Foundation to conduct (with Tassinary) advanced training workshops in social psychophysiology. Workshop participants, who were mostly established researchers in social psychology, came to the University of Iowa (and later the Ohio State University) for four weeks of intensive training. Graduate and undergraduate students in John’s lab served as teaching assistants each 8-hour day, in which every aspect of social psychophysiology was covered, including a comprehensive introduction to nearly all psychophysiological methods (e.g., facial EMG, EEG, electrodermal activity, cardiovascular measures), experimental design, and advanced analytic techniques. Guest lecturers were flown in to give state-of-the-art reviews in their area of expertise, and their lectures and demonstrations were video-recorded for viewing at later summer schools. In total, there were six summer schools from 1986 to 1990, and among the notable participants were Gordon Bower, Nancy Eisenberg, Russell Geen, Gerald Ginsberg, John Gottman, Anthony Greenwald, Elaine Hatfield, Bill Ickes, Randy Larsen, Norman Miller, Cookie Stephan, Walter Stephan, Jerry Suls, Abe Tesser, Dianne Tice and many others. A primary goal of these workshops was to train leaders in the field enough about social psychophysiology so that they would be knowledgeable about it and would hence be more amenable to its methods when reviewing or making decisions about funding, etc. Another goal, however, was to enhance the skills of the researchers themselves so that they would start using such methods in their own research. Many of them did, in turn training the next generation of social psychophysiologists.

Importance of sound inference

One of John’s enduring contributions has been the series of a papers that he coauthored on psychological inference. A starting point for this work goes back to a pair of papers (Cacioppo et al., 1983; Dorfman & Cacioppo, 1990) that are, strangely enough, among the least cited of his more than 500 scientific publications. These two papers introduced to psychophysiologists the use of waveform moment analysis, originally developed by mathematicians, to apply to the interpretation of bioelectric signals (e.g., EEG, EMG) in the amplitude, time, and frequency domains. John and his colleagues argued that too much important information was being lost by simply, for example, computing the mean amplitude of the voltage of a signal as an index of the psychological construct of interest. Although this analytical approach was never widely adopted by psychophysiologists, the basic problems of inferring psychological meaning from physiological activity became a major focus of his writings from that point on.

John was at heart a psychophysiologist. Psychophysiology, since its formal establishment as a discipline in the mid 20^th century, via the journal Psychophysiology and the Society for Psychophysiological Research, regularly deals with the challenge of measuring physiological activity and using the obtained data to draw inferences regarding psychological processes. Of course, there is no direct mapping of any physiological activity to a single psychological process. With a clear influence of William James, John wrote much over the years about the relationships between processes in the physiological and the psychological domain and established clear typologies of relationships between them.

For example, Cacioppo and Tassinary (1990) presented a taxonomy of relationships between psychological and physiological events, of which generality (i.e., the extent to whichthe relationship is context-dependent or context-free) and specificity (the extent to which the relationship is one-to-one or many-to-one) are the two primary dimensions. This taxonomy was later updated (Cacioppo et al., 2007) to include a third dimension, sensitivity, which is the degree to which changes in the physiological measure reflect changes in the underlying psychological construct. In all, the taxonomy has clear relevance to a persistent challenge to all of social neuroscience – how much can one really infer the psychological significance of a physiological event? Interested readers are strongly encouraged to read these works. The clarity and logic apply equally well to linking personality tests to psychological processes and in making sense of the activations observed in fMRI, as reflected in current conversations about reverse inference. Of course, John was not the first to reflect and write on these matters, nor will he be the last. However, his clarity, style, and logic in addressing the issue of inference will be difficult to match.

Another example of John’s interest in inference issues is a paper he published with his colleagues (Cacioppo et al., 2003) cleverly titled “Just because you’re imaging the brain doesn’t mean you can stop using your head: A primer and set of first principles.” This paper was published in Journal of Personality and Social Psychology at a time when articles authored by social psychologists using neuroimaging were proliferating. Many of these papers, although they employed the latest neuroimaging techniques, faltered when the validity of their designs and methods was more thoroughly considered. John and his colleagues offered four principles for those new to social neuroscience to consider:

1. (We know already that) social cognition, emotion, and behavior involve the brain.

2. The functional localization of component social processes or representations is not a search for “centers.”

3. Localized changes in brain activation that differ as a function of a task do not, in themselves, signal a neural substrate.

4. The beauty of a brain image does not speak to the psychological significance of the image.

Of course, the underlying inferential logic in this paper could already be found in his earlier writings, but now it was being taught to a new generation of scientists with powerful MRI scanners.

Defining social neuroscience

This leads us, of course, to perhaps John’s most noted milestone with respect to this journal and this special issue – the co-founding of the field of social neuroscience. As noted, prior to the 1990s, the work that John and others were doing was labeled “social psychophysiology,” and it was quite comprehensive in both its methodology and theoretical approach. Recently, one of the present authors (EJV) attended a tribute to John in which a speaker argued that, because John did not have a way to image the brain in those early days, he had to rely on peripheral psychophysiological measures to do his research. This seems to be far from true. In their first use of the term “social neuroscience,” Cacioppo and Berntson (1992) made a convincing argument that social psychologists/psychophysiologists had a lot to offer the field of neuroscience. In this same paper they introduced their doctrine of multilevel analysis, and argued that social psychologists and neuroscientists could benefit from a consideration of the other’s research. In this paper and in John’s presidential address to the Society for Psychophysiological Research (Cacioppo, 1994), John’s and his colleagues’ latest research on the relationship between cardiovascular, neuroendocrine, and immune responses was presented as a powerful example of this new social neuroscience. Social factors, such as social isolation or caring for a terminally ill family member, were identified as having important effects on these relationships. Although he and his colleagues would soon begin an important line of research in EEG and ERPs, in 1993 John clearly envisioned social neuroscience as having multiple levels of explanation and requiring multiple kinds of physiological evidence. It was not all just the brain!

Nearly a decade later, John (Cacioppo, 2002) elaborated on his vision of social neuroscience, celebrating the collaboration of neuroscientists and cognitive scientists in the previous decade, and now calling on more collaboration of social scientists and neuroscientists. In 2006, this journal, Social Neuroscience, was founded, as was Social Cognitive and Affective Neuroscience . The Social and Affective Neuroscience Society was founded in 2008. Its focus is on the neural basis of (human) social and affective processes. Two years later, the editor of Social Neuroscience, Jean Decety, and John traveled together and consulted with many like-minded researchers around the globe, and ultimately decided to the launch the Society for Social Neuroscience in 2010. On its website, the society defined social neuroscience as “the interdisciplinary study of the neural, hormonal, cellular, and genetic mechanisms underlying the emergent structures that define social species” (Society for Social Neuroscience, n.d.).

Legacy

When planning this special issue in celebration of the work of John Cacioppo, we deliberately invited contributors whom we believed well represented the different approachesto social neuroscience that John himself championed. We are thrilled that this group of outstanding authors accepted our invitation and completed their papers on time. Nine papers were submitted and went through peer review. Eight appear in this issue. Due to a publishing error, the ninth paper has already been published (Knoll et al., 2020), but should be considered as part of this collection in honor of John.

Disclosure statement

No potential conflict of interest was reported by the authors.