Qualitative Methods

Traditionally, emotion researchers use a wide range of qualitative methods, most consisting of verbal tools like questionnaires and very often based on self-assessment. They cover a variety of topics ranging from quality of life, empathy, and hedonism to emotional intelligence. More specific instruments have been developed, including structured interviews for typical clinical groups such as the Autism Diagnostic Observation Schedule (ADOS)for autism (Lord et al., 1999) or the Positive and Negative Syndrome Scale (PANSS) for schizophrenia (Kay et al., 1987). Interestingly, in the last decade we have witnessed many efforts to relate the products of these qualitative measures to sophisticated analysis of the results obtained in functional brain-imaging studies. The notion is that they may actually validate each other. Typically, the argument runs that quantitative measures provide the objective underpinnings of the qualitative measurements.

Quantitative Methods

Obviously, the overall goal of quantitative measuring methods is to develop a metric that allows measuring the relation between the emotional experience, the behavior, and the neural events that give rise to it and are associated with it. In 1890, William James already identified the challenge in two chapters of his Principles of Psychology. Most succinctly, and in words that ring as true now as then, James notes, “A science of the mind must reduce complexities of behavior to their elements. A science of the brain must point out the functions of its elements. A science of the relations of mind and brain must show how the elementary ingredients of the former correspond to the elementary ingredients of the latter” (James, 1890/1950, p. 28). A striking aspect of the problem, as James already described it, is that with the advent of more and more measuring methods, the task has become even more difficult. What are the elements James is referring to? How do we individuate elements of the mind, how do we individuate functional atoms in the brain, and how do we relate each to the other? The problem is, in a nutshell, that with every new method that becomes available to measure the mind-brain and specifically its emotions, a new way of individuating elements becomes available. New ways of carving up the mind, and new ways of carving up the brain, make for even more new ways to conceive of the relations.

A brief review of the most commonly used methods confirms that this is indeed the case and that each method has its own ways of defining neurofunctional bases and temporal dynamics of mind-brain processes. While one cannot measure emotions only with self-assessment questionnaires or only with objective psychophysics or brain imaging, each approach does measure some aspects of emotions. In the following paragraphs we provide a few selected examples that are particularly relevant for the topic.


Ever since Fechner, empirical methods have been focused on establishing what the interface is between physical quantities and subjective experiences. Many treatises have been devoted to the foundations, the analysis, the interpretation, and the best use of psychophysics methods in behavioral experiments. A century of psychophysical studies represents a treasure of scientific thinking on the human mind. Currently the sophistication of this knowledge is unsurpassed. The combination of psychophysical methods and the neurofunctional methods to be mentioned now promises significant progress.


The neuropsychological method was traditionally a major source of new insights and theories because it allowed making inferences from behavioral patterns to underlying brain disorders. A prime example of this is the set of studies on the memory impairments of HM. Patient HM significantly influenced theoretical developments in memory research. Rare patients continue to play a significant role in affective neuroscience, much as they have done in other areas. Among the most famous cases is Damasio’s patient XX. The study of populations selected for a clear behavioral and/or neurological deficit, and of unique deficits seen in single cases, continues to be a very important source of insight into affective processes.

Other insights continue to come from groups of patients. Populations that are of special interest for the study of emotion are those with frontotemporal dementia, autism (see Maurer & Damasio, 1982), schizophrenia (e.g., Arnold et al., 1991), and obsessive-compulsive disorder as well as Alzheimer (e.g. Hyman, Van Hoesen, Damasio, & Barnes, 1984) or Parkinson patients (Damasio, 1979). There is increasing awareness that emotional life and, for that matter, the brain’s emotional systems are not isolated nor cut-off from cognition. Cognitive disorders erode emotional well-being. By investigating some of these disorders using bodily expressions as stimuli, we avail ourselves of a new window into these clinical symptoms and hopefully will be able to throw a new light on them.

functional brain imaging

Within the last decade, brain imaging, mainly using fMRI, has been the method of choice for many researchers. Recently, emotion researchers following this basic emotion category approach have attempted to pin down the brain representation of a few principal emotions. For example, fear was associated with the amygdala (e.g., Morris et al., 1996), disgust with the insula (e.g., Phillips et al., 1997), and so forth. After what looked like a promising beginning, a more complex situation emerged. To stay with the example of the amygdala, follow-up reports showed that anger expressions also activated this structure (e.g., Whalen et al., 2001) and that, depending on the design of the study and the stimuli used in the specific study, neutral faces also clearly triggered amygdala activation (e.g., Morris et al., 1998; Todorov & Engell, 2008).

In line with this Lindquist et al. (2012) are rightfully critical of the fact that meta-analyses inherit the weak points of still less-than-perfect brainimaging techniques and cannot but endorse and amplify them. Functional MRI studies vary widely in scanner properties, settings, designs, and tasks, including the involvement of attention, awareness, and contrast stimuli or conditions. The meta-analysis exploits the very procedures under attack by using positive activation levels of isolated brain areas, themselves obtained in a wide variety of studies. The meta-analytic conclusion that some area may play or not play its anticipated role does not invalidate that role, and this role may or may not show up in fMRI analysis. For example, the amygdala was repeatedly shown to play a role in processing of emotional stimuli, and brain-imaging studies of autism are consistent with this (e.g., Baron-Cohen et al., 1999; Pierce et al., 2001). However, patients with Urbach-Wiethe syndrome have a major deficit of the basolateral amygdala, yet show no signs of autistic behavior (Paul et al., 2010). In fact, they are very alert to social signals, and their gaze pattern is no different from those of neurotypical controls (unlike what is found for autistic subjects). There are many more examples illustrating this lack of a rigid link between a brain area and a functional deficit. But the suggestion of attributing functions to a network rather that to a single area may also beg the question. One would like to think that understanding the network implementation and understanding the theory must go hand in hand.

< Prev   CONTENTS   Source   Next >