Abstract: It has been said that mirror neurons are “the most hyped concept in neuroscience”. There are two main and robust hypotheses that have been advanced on what might be the functional role of mirror neurons. One suggests that mirror neurons are at the basis of action understanding while the other emphasize that mirror neuron activity mediates imitation. However, as many researchers have recently discussed, the term “action understanding” has not been clearly defined. It has been used to refer to action perception, action recognition and selection. According to the “direct matching hypothesis” designated by Rizzolatti and colleagues, an action is understood when its observation makes the motor system of the observer “resonate”. Thus, when we observe a man biting an apple, the same population of neurons that control the execution of biting will resonate in the observer’s motor areas. A more radical definition for action understanding is a deflationary account proposed by Gallese, “determining why a given act (e.g. grasping a cup) was executed can be equivalent to detecting the goal of the still not executed and impending subsequent act (e.g. bringing the cup to the mouth). ” There are constitutive and causal interpretations of mirror neurons’ role in action understanding, but unfortunately, they are rarely distinguished in the current literatures. To say that neural phenomenon λ constitutes cognitive phenomenon ζ is like saying that λ realizes ζ, but this claim is potentially circular or begging the question in standard philosophical wisdom. The causal relation stands in contrast to the constitution relation. To say that event λ causes event ζ is to say that λ brings about ζ. There are three kinds of causal relations, necessary cause means that λ is necessary to bring about ζ; sufficient cause means that λ is sufficient to bring about ζ; and contributory cause means that λ is neither necessary nor sufficient to bring about ζ, but nevertheless it contributes to bringing about ζ. Recently, a series of empirical evidence supports the idea that mirror neurons are contributory causes of action understanding. In order to clarify how mirror neurons contributory cause action understanding, it is very necessary and to profitably distinguish between movements, motor acts and motor actions from both a conceptual and a neurophysiological points of view. A movement is defined as the displacement of joints or body parts without a specific goal (that can be observed both in everyday activities or can be induced by electrical or magnetic stimulation of the motor area). However, recently a bit of controversy arose over the mirror neuron activation for movement. A motor act directly toward an object or a body and potentially leads to an interaction, which encodes a specific object-directed interaction with a proximal goal (e.g. grasping a chip). Evidence on non-human primates has demonstrated that a motor act is coded by the activation of strictly and broadly congruent mirror neurons. A motor action is defined as a sequence of motor acts directed towards a distal goal (e.g. reaching for a piece of food, grasping it, holding it and bringing it to the mouth). Current work that combines behavioral studies with neuroimaging and transcranial magnetic stimulation in human indicates that mirror neuron activity, in conjunction with the perceptual cues, causes one to recognize and anticipate the completion of a goal-directed action. This shows action understanding is mediated by an inferential system (e.g. STS and mPFC, etc.). Notably, for proponents and opponents of deflationary account, to compromise a compatible approach, like different versions of dual process theories of action understanding, may be a relatively easy choice. On the contrary, direct responding and defending to the challenge from each other must the first important work to do.

Key words: Action Understanding, Mirror Neurons, Contributory Cause, Deflationary Account, Goal

摘要： “动作理解”一直被视为镜像神经元功能最为保守的假设，同时也是证据链最为强健的领域之一，但其概念内涵却始终如同古希腊神话中的“戈尔迪之结”一般没有得到厘清。目前，动作理解的紧缩解释正在面临严峻的挑战。来自实验室检验与哲学拷问的证据倾向于认为，镜像神经元活动构成动作理解的观点是一种“循环论证”，前者至多只能算作产生动作理解的附带原因，而不是必要与充分原因。将广义的动作分成运动、运动动作为与运动动作链可以有效地澄清实现理解不同层次的动作需要具备哪些条件。回应质疑并为辩护自身立场指明前进方向，是重解动作理解“戈尔迪之结”的起点，将对镜像神经元的科学祛魅产生奠基性的影响。

关键词: 动作理解, 镜像神经元, 紧缩解释, 附带原因, 目标