A generic technique for measuring facial action, the FACS was developed in the mid-1970s and published in 1978 by Paul Ekman and Wallace Friesen. The FACS was designed to allow a comprehensive description of any action of which the face is capable. One of a number of techniques that allow investigators to characterize facial movements, the FACS was developed to overcome some of the conceptual and methodological difficulties in previous systems, chiefly the lack of independence between a description of an event observed and the conceptual premises of the observer. Historically, systems for measuring facial action have been inferential, requiring observers to interpret the event seen before describing it. In addition, they have often required the observer to describe facial acts in terms of configurations of actions that represent gestalt judgments. For example, and observer might be required to decide whether a facial action represents a smile, a frown, or some other global action to which a meaning is imputed. Such systems have proved to be problematic for several reasons: different observers hold different criteria for making judgments of the same event; observers' definitions of the behavior in question change over time; and many instances of facial behavior do not correspond to the categories designated in the system yet still have meaning.
The FACS overcomes some of these problems. It is based on an anatomic-measurement principle and describes a facial action in terms of the underlying muscular action that produces it. An observer making use of the FACS will “dissect” a facial action into its muscular bases. The system contains rules and decision criteria that allow the discrimination of forty-four separate actions, most of which can be present independently on either side of the face.
The FACS contains descriptions of action units (Aus), which are the fundamental actions of individual muscles (e.g., the zygomatic major, which produces the lip-pulling action characteristic of a smile) or groups of muscles that characteristically act in unison (e.g., the depressor glabellae, the depressor supercilii, and the corrugator that together draw the eyebrows in a downward and medial direction), and action descriptors (Ads), which are unitary movements that may involve the actions of several muscle groups (e.g., a forward-thrusting movement of the jaw). Each AU is labeled by a number and a simple description of its action. This approach helps preserve the distinction between actions and their interpretation and discourages observers from imposing meaning upon the actions.
FACS coding is ordinarily performed using a videotaped record of the action at issue, although coding from still photographs is possible. To perform FACS coding, the observer views the action repeatedly in real time and in slow motion, applying the decision rules to come up with an ultimate code. Various measurable parameters of facial actions include time of onset, time of offset, time of apex, rise time, decay time, and, for many of the actions, intensity. Because it requires an observer to make repeated observations and decisions about a large number of codes, it is a more laborious and time-intensive procedure than many other behavior-measurement systems. Just how laborious depends in part on the density of actions that are taking place; however, a ratio of coding time to real time of 100:1 is typical. This ratio can be reduced if the investigator is interested in only a limited set of facial actions. For investigators who are interested in actions that are thought to be involved in emotion, an abbreviated version of the system, EMFACS has been developed. In addition, a computer program (the emotion dictionary) is available that enables deductions about emotional expressions from raw FACS coding.
Encyclopedia of Semiotics. © 1998
Oxford University Press.
All rights reserved.
