Approaches to pain assessment typically distinguish between self-report and nonverbal, observational indices, but do not specify criteria or reasoning for the distinction (Hadjistavropoulos & Craig, 2002). Self-report indicators usually employ verbal report to describe the individual’s pain experience, or to communicate qualities of experience through the use of speech (e.g. ‘I have a dull nagging pain in my lumbar region’). The context is important as the exchange is either initiated by the individual in pain or serves as a response to others’ queries. Observational indicators typically focus upon nonverbal, publicly observable manifestations of pain including facial expressions, cry, limb and torso movements and paralinguistic qualities of speech.

The nonverbal behavior that is the focus of observational indices typically can be viewed as primitive in the sense that the behavior represents reflexive patterns of response to actual or impending tissue damage that are not anteceded by introspective awareness or inner observation of private phenomena (Wittgenstein, 1958b; 1980b; 1982). For example, babies’ cries typically alert mothers and caretakers as to children’s needs and initiate required care, even though the infant cannot be characterized as intending or consciously understanding the interaction. While pain is a powerful initiator of attention, introspection does not mediate or modulate primitive behavioral reactions. Instead, we claim that when conscious awareness is employed to influence any form of social communication, the expression cannot be taken as an automatic reaction pattern. For example, to pretend to be in pain, an individual must know how a person who is in pain behaves and focus one’s behavior on that model, intend to reproduce it (Wittgenstein, 1982). Moreover, people tend to experience difficulty suppressing nonverbal expressions of pain and there are subtle differences between spontaneous and feigned expressions of pain (Craig et. al, 1999).

Another important aspect of the primitiveness of pain communication pertains to the experience of conscious awareness of what is happening. Self-report indices typically necessitate attention to the task, including some degree of self-observation. By contrast, the behaviors included in many nonverbal indices occur spontaneously without prior awareness, although the individual likely will monitor the display (Hadjistavropoulos & Craig, 2002). This view ignores the important fact that self-report is used as a partial substitute for automatic nonverbal behaviors (Wittgenstein, 1958b). The exclamation ‘It hurts’, the groan ‘I have a stomach-ache’, the cry ‘I’ve hurt myself’ are reflexive patterns of response, not descriptions; they are comparable to automatic nonverbal behaviors rather than to reports such as ‘He has a back-pain’. In this situation, there is no such thing as my ‘finding out’ that I am in pain, or of my ‘attending’ or ‘recognizing’ my pain-sensations or experiences (Wittgenstein, 1958b).

A primitive expression of pain can be over-ridden to some degree. For instance, facial pain expression can be voluntarily controlled, although the upper face is less amenable to conscious control (Rinn, 1984). Communicative acts often reflect anticipation of consequences. As Wittgenstein (1958b) noted, pain behavior is deeply influenced by conditioning and learning. We claim that the more dependent pain behavior is on learning and conditioning, the more it will be subject to deliberate mediation. Empirical evidence suggests that nonverbal pain expressions that are genuine can be distinguished from masked or exaggerated expressions (Hadjistavropoulos et. al, 1996), possibly because of the greater automaticity of nonverbal expression places limits on the degree to which it can be influenced by learning and conditioning (in contrast to self-report). Observers generally appreciate this and communicate preferences for nonverbal behavior over verbal behavior when assessing the truthfulness of others’ pain behavior (Craig, Prkachin, & Grunau, 1992). This suggests that pain assessment presupposes behavioral manifestations, primarily nonverbal and observational indices, and that ascribing pain in others is possible only where we have criteria for identifying pain, which means that pain must be capable of being expressed (Wittgenstein, 1958b).