ABSTRACT
This article will analyze the findings of Iacoboni et al. (2005), wherein it is stated, “The human mirror neuron system does not simply provide an action recognition mechanism, but also constitutes a neural system for coding the intentions of others” and will suggest that Iacoboni et al.’s findings are not supported by the results of their experiments.
Introduction
Iacoboni et al. (2005) tried to determine the difference in mirror neurons activity in the brain when simply observing an action with no context (e.g., grasping a cup) and observing the same action within a context (e.g., drinking or cleaning). The difference in response would indicate that the additional elevation of the signal, above the Action and Context alone, can be attributed to the understanding intention behind an Action or “what comes next?” in intention conditions.
To test this competing hypothesis, Iacoboni et al. studied normal volunteers using functional magnetic resonance imaging, which allows in vivo monitoring of brain activity. Twenty-three subjects in the experiments watched three kinds of visual stimuli, introduced as (a) Action, grasping hand actions with no context; (b) context (containing only objects), drinking or cleaning; and (c) two conditions, Intention drinking and Intention cleaning, the grasping actions embedded in two different contexts (either drinking or cleaning) (see Fig.1).
It was found that observing grasping actions embedded in contexts yielded greater activity in mirror neuron areas in the inferior frontal cortex than did grasping actions in the absence of contexts or while observing contexts only. These results led to a conclusion that “the human mirror neuron system does not simply provide an action recognition mechanism, but also constitutes a neural system for coding the intentions of others” [1]
We will demonstrate that these findings are based on incorrect premises, leading to an improper interpretation of the findings. We will then offer an alternative explanation for the high signal increase in the inferior frontal cortex for Intention drinking and for the differences between the two intention
conditions.
conditions.
Discussion of the Central Findings of Iacoboni et al., 2005
One aspect of the investigation by Iacoboni et al. (2005) was to measure and compare the activity levels among conditions Intention drinking, Intention cleaning, Action, and Contexts in areas of the human brain known to have mirror neuron properties. The results of the experiments showed that Intention cleaning produced a response similar to Context cleaning (see Fig. 4). Intention cleaning produced a response slightly higher than Action alone. The magnitude of the signal in Intention cleaning can be attributed to Context cleaning, as the sum of Action plus Context cleaning, leaving no extra signal “over the peak of cleaning Context,” to attribute to the understanding of intention. In addition, any slight discrepancies between these two can be ignored on the basis of the principle invoked by Iacoboni et al. when they stated that the difference between drinking and cleaning Contexts (which is actually in double digits compared with the difference in response between Intention cleaning and Action) offers “no reliable difference between the ‘drinking’ Context clip and the ‘cleaning’ Context clip (p > 0.19) [1].
The lack of any significant increase in signal in the Intention cleaning condition suggests that one cannot generalize that “observing grasping actions embedded in contexts yielded greater activity in mirror neuron areas in the inferior frontal cortex than did grasping actions in the absence of contexts or while observing contexts only”[1]. This last statement is clearly not supported in the Intention cleaning condition by the graphically displayed results as shown in Figure 4, below.
Fig. 4
Furthermore, Figure 3 does not represent the Intention cleaning condition at all to support the general statement that “the Intention condition yielded significant signal increases…in the inferior frontal gyrus” [1].
Fig. 3
From the second area of investigation, the difference between Intention drinking and Intention cleaning conditions shows that “the “drinking” Intention clip yielded a much stronger response than the “cleaning” Intention clip (p < 0.003), see Fig. 4. This difference was the underpinning for the discussion aimed at validating the intention coding hypothesis. The following line of arguments was offered to support the hypothesis:
If grasping actions in the intention conditions are identical (produce the same mirror neuron activity in the inferior frontal cortex), and the corresponding Contexts alone produce the same levels of mirror neuron activity, then the increase of the signal in intention conditions (action embedded in a context) should be the same. If the increase of the signal is different, it implies that this area actively participates in understanding the intentions behind the observed actions.
“If the grasping action embedded in two different contexts (drinking and cleaning) are identical and corresponding Contexts alone produces the same mirror neuron activity (no reliable difference was observed between the ‘drinking’ Context clip and the ‘cleaning’ Context clip (p > 0.19), then the increase of the signal should be the same in both intention conditions, Intention drinking and Intention cleaning” [1] In short, if (a) and (b) are the same (actions), (c) and (d) are the same (contexts), then the combination between (a) + (c) and (b) + (d) (intentions) should produce the same results. And if the results are different, this could indicate that an additional “unknown” factor might influence the results. The results of the experiments demonstrate that the level of activity between both conditions [(a) + (c) and (b) + (d)] are significantly different (see Fig. 4), and these differences, according to Iacoboni et al., can be attributed to the intention.
The preceding reasoning would definitely be applicable and supportive if the grasping actions in all three conditions were identical. However, if we look closely at Figure 1, which depicts clips as the actual experiments were performed, one can easily determine that the grasping actions in all clips are not identical. For example, in Action, the grasping action consists of the hand grasping an empty cup, while in Intention drinking, the hand is grasping a cup filled with tea. In Intention cleaning, the hand is grasping an empty cup again. This suggests that the participants observed different grasping actions, which vary not only biomechanically but also by the differences in the type of main objects.
To advance this analysis, it is first necessary to introduce two new operational terms: experimental model and postulated model. The experimental model is what is observed by the participants and examined by the investigators (see Fig. 1). From this model, actual results are derived and conclusions are formulated. The postulated model is a propositional model (see Fig. 6a), which serves as the basis for the arguments and conclusions.
The postulated model consists of the three Action conditions with identical grasping actions (see Fig. 6a), while the experimental model consists of the three Action conditions with different grasping actions (see Fig. 1). For Iacoboni et al. to justify the findings, the two should be the same: “These findings clearly show that coding intention activates a specific set of inferior frontal cortex neurons and that this activation cannot be attributed either to the grasping action (identical in both ‘drinking’ and ‘cleaning’ Intention clips) or to the surrounding objects, given that these objects produced identical signal increase in the ‘drinking’ and ‘cleaning’ Context clips” [1]
Figure 6a
Hypothetical model: identical grasping actions in Action, Intention drinking, and Intention cleaning clips.
On the one hand, we see the experimental model, Figure1, with different grasping actions; and on the other hand, we referred to the postulated model with identical grasping actions (see Fig 6a). By considering this distinction, it is both plausible and logical to suggest that the varying responses in the intention conditions may be causally related to the observation of different types of entities. To support this proposition, we present new figures, numbered 6a through 6e (sequentially following Iacoboni et al.’s figures 1 through 5 in the original text).
Alternative Explanation to the Experimental Data
Identical Grasping Actions
Iacoboni et al. (2005) examined three action conditions, which they labeled Action, Intention drinking, and Intention cleaning (see Fig.1). Each of the three conditions were purported to consist of an identical grasping action. An example of the grasping action is given in the middle clip, “Action,” of Figure1. To clarify the argument, we will rename “Action” as A1 (the hand grasping an empty cup). Action in the Intention drinking clip we will rename as Ä2 (the hand grasping a cup filled with tea). Action in the Intention cleaning clip we will rename as A1 (the hand grasping an empty cup). The new labeling scheme is illustrated in Figure 6b, which is derived from Figure 1, the experimental model.
A1 A2
A1 A1
Figure 6b
(A1 for Action clip, A2 for Intention drinking clip, and A1 for Intention cleaning clip)
These differences in the type of graspable objects and the type of grasping actions are recognized by Harmon-Jones (2007), who states, “when the cup [empty in one case and food in the other] is going to be grasped by somebody, it is no longer simply a cup; it belongs to the grasping action” [2]. Grasping actions cannot be considered actions without particular graspable object(s) with particular physical properties. In the first case, Action, the object is an empty cup, whereas in the other case, Intention drinking, the object is essentially two objects: the food and the cup. Both are being visually grasped. This accords with the results published by Di Pellegrino (1992), who concluded that object(s) of grasping in grasping actions most likely influence the mirror neuron activity level. “The effective experimenters’ movements [the higher response of the mirror neurons’ activity] include among others placing or retrieving a piece of food from a table, grasping food from another experimenter’s hand,” compared to non-food responses[3]. This finding is supported by Gallese (1996), who reported that mirror neurons always discharge to food-related actions, while the response diminishes or disappears “after a few or even the first presentation” to non-food three-dimensional solids[4]. .
Twenty-three subjects watched three kinds of visual stimuli in Iacoboni et al.’s experiments. The participants observed and analyzed the differences between the objects (food as well as non-food) by their various physical properties (color, shape, volume, texture, location, etc.). Most likely, picking up food will elicit a very different level of neural activity, compared to picking up empty cups or plates. Therefore, comparing the amount, location, or intensity of neural activity between these two cases may be misleading. It follows logically that the participants in Iacoboni et al.’s study did not observe identical or interchangeable sets of entities, as the researchers purport. The most plausible explanation for the different levels of neural activity in the inferior frontal cortex for intention conditions may lay within the differences of observed entities in Figure 1.
The following sections of this paper will offer an explicit explanation to support our proposition.
Grasping the intentions of others with one’s own mirror neuron system
An important argument in support of the intention coding hypothesis “Grasping the intentions of others with one’s own mirror neuron system” is offered by Iacoboni et al. (2005). The researchers proposed that the understanding of intention is derived from the observation of an action embedded in a context, thus, suggesting the following action after the observed one. The following action, in turn, provokes the additional activation in the mirror neurons area. This additional activation manifests the understanding of intention.
Iacoboni et al. ascribed a significant role in their research to contexts surrounding actions, where they emphasized the importance of identity between grasping actions in the Intention drinking and Intention cleaning. “An important clue for clarifying the intentions behind the actions of others is given by the context in which these actions are performed. The same action done in two different contexts acquires different meanings and may reflect two different intentions.”[1].
To prove their case, Iacoboni et al. offered an example of a following action, functionally related to the observed one, in the Intention drinking condition: the action of “the bringing to the mouth.” (Note that in the Intention cleaning condition, they do not suggest a following action, functionally related to the observed one.)
According to Iacoboni et al.’s experimental model (Fig. 1), the Intention drinking condition, “bringing to the mouth” action, as suggested by a context, would follow after A2 . Both actions A2 and “bringing to the mouth” are food-related actions. However, A2 is not the same as A1.
This inconsistency contradicts the logic of Iacoboni et al.’s argument that “the same action done in two different contexts acquires different meanings and may reflect two different intentions,” [1] and, therefore, the whole argument losses its credibility. One can argue that these inconsistencies are insignificant and will not affect the line of logic linking context, intention and the following action, suggested by a context. If this is the case, let us consider actions A1 (hand grasping an empty cup with one type of grip) and A2 (hand grasping a cup filled with tea with another type of grip) as identical and interchangeable, as Iacoboni et al. in essence purport (see Fig. 6c). Then this alleged equality between A1 and A2 would allow us to exchange components within the conditions of the experimental model, Figure 1: A1 placed in the drinking context and A2 placed in the cleaning context, as shown in Figure 6d.
Figure 6c
(A1: the hand grasping an empty cup; A2: the hand grasping a cup filled with tea)
Figure 6d Imaginary Model
In case of A1 (hand grasping an empty cup) in a drinking context scenario, the “bringing to the mouth” action does not follow, (Fig. 6d) while a drinking context suggesting “bringing to the mouth” as the following action. Here we find a contradiction with the applied logic of Iacoboni’s argument that “the context cued the intention behind the action.”. Conversely, were the hand to grasp a cup filled with tea, A2, placed in the cleaning context “bringing to the mouth” action would be applicable, but the context itself does not suggest “bringing to the mouth” as the following action. Context in both versions creates confusion in the understanding of the following action, or what the agent will likely do next. This functional incompatibility between contexts and the base actions contradicts Iacoboni ‘s broad general assertion: “An important clue for clarifying the intentions behind the actions of others is given by the context in which these actions are performed.”
The following important points can be drawn from the foregoing analysis: 1. Confusion in predicting the following action from the context is tantamount to confusion in the understanding of intention. This, in turn might prevent and preclude the additional activation in the mirror neuron area of the participants. “The most straightforward interpretation of our results is that the selection of these neurons is due to the observation of an action, ... in a context in which that action is typically followed by a subsequent specific motor act.” Therefore, the main argument to support hypothesis loses its creditability. 2. Confusion in predicting the following action may suggest that these actions A1 and A2 cannot be considered as identical and, therefore, that implies the type of observed entities “grasping actions”, plays a significant role in modulation of activity in mirror neuron areas.
The Role of Context in the Experimental Model
The operational definition of the word context seems to vary from clip to clip. For example, the upper clip of Figure 1, Context Before Tea (or drinking context) appears to consist of edible items, including a cup with tea. In the Intention drinking clip, Action consists of main object, cup with tea, while this object is a part of the drinking context. The empty cup is missing in the Intention drinking clip, and the cup with tea is missing from the drinking context.
Similarly, the Context “After Tea” appears to consist of inedible items, including an empty cup. Therefore, the empty cup is a component of the context in the given clip, and it is the main object of the Intention cleaning clip.
The consequences of inconsistency in experiments can be illustrated in the lower row of Figure 3, “Signal Increases for Intention Minus Action and Intention Minus Context.” In Iacoboni et al.’s study, Intention Minus Context must result in Action. However, in Intention cleaning Minus Context cleaning this would result in a scene showing only a hand movement, mimicking motor act. There will be no object to which a motor act is directed. Most likely, no mirror neuron activity will be exhibited. In fact, Intention cleaning minus Action has no representation at all in Figure 3.
The role of a context in the given experiments can be demonstrated by comparing the levels of the mirror neuron activity between conditions that employ the same grasping actions. This applies to Intention cleaning and Action, A1. As shown in Figure 4, the difference in the mirror neuron response between Intention cleaning and Action is minimal, essentially equal to zero. Therefore, the results may suggest that context did not influence activity in mirror neuron areas, thus implying that the mirror neuron system simply codes the type of observed action. This interpretation is directly supported by Iacoboni et al.: “If the mirror neuron system simply codes the type of observed action and its immediate goal, then the activity in mirror neuron areas should not be influenced by the presence or the absence of context” [1].
The role of context in the Intention drinking of the postulated scenario (base action A1, empty cup in all three action conditions; Fig. 6a) cannot be determined, because the level of the mirror neuron activity in Intention drinking with base action A1 was not measured.
The role of context in the Intention drinking condition in the experimental scenario cannot be determined, since the level of A2, as a single action, was not measured to compare A2 with Intention drinking, both of which employ the same grasping actions, A2. These conditions would be a part of a plausible scenario shown in Figure 6e, where each of three Action conditions employs the same grasping action (full cup only), A2. To validate our position, the level of neural activity in Action (full cup only) can be measured, Ä2, to find if we get the same level of activation as in Intention drinking, as shown in Iacoboni et al.’s Figure 4.
Fig. 6e: Action, Intention drinking and Intention cleaning conditions with identical grasping actions, A2.
Three Phases of a Grasping Action
Under Gallese’s (1999) definition, a grasping action consists of two phases: “First, is the ‘pre-movement’ phase, during which the physical properties of the object to be grasped, like its size, its shape, its axis orientation, are analyzed. The second phase is ‘an executive phase,’ during which the subject’s hand (in this case, a monkey) is aimed at the object in such a way as to properly match its previously analyzed intrinsic properties” [5]. Bertenthal (2005) suggested that “when we observe a hand grasping an apple, the analyzed elements would be the hand, the apple, and the movement of the hand toward the apple” [6]. Iacoboni et al. (2005) clarified the essential component of grasping actions: “John sees Mary grasping an apple. By seeing her hand moving toward the apple….” To conduct the experiments, Iacoboni et al. introduced a model of the grasping action, the middle clip of Figure 1, which illustrates that the object (an empty cup) is actually grasped, being in the hand and eyes of the performer.
By integrating the collective thoughts of the researchers cited above, we can conclude that a grasping action consists of the three phases. The first phase, as Gallese (1999) notes, is the “premovement phase, during which the physical properties of the object to be grasped, like its size, its shape, its axis orientation, are analyzed.”
The second phase involves an actual hand movement toward the apple (which accords with Bertenthal (2005) and Iacoboni et al., from the performer’s body (point A) to a desired object—for example, an apple (point B). Think of this as the performer’s reaching phase, “during which the performer’s hand is aimed at the object in such a way as to properly match its previously analyzed intrinsic properties” Gallese [5].
The third phase is actually possessing, holding, and feeling the object. This phase is the one that is seen by the participants who watched the Action, Intention drinking, and Intention cleaning clips in Figure 1.
From Action Recognition to the Coding of Intentions
Iacoboni et al. reported that the results of their experiments suggest that the role of the mirror neuron system extends from action recognition to the coding of intentions: “The data of the present study suggest that the role of the mirror neuron system in coding actions is more complex than previously shown and extends from action recognition to the coding of intentions” [1]. This data comes from the comparison of the signal level in the inferior frontal cortex among three Action conditions, and particularly between the two intentions conditions: “the ‘drinking’ Intention, which yielded the stronger response, and the ‘cleaning’ Intention clip (p < 0.003; Fig. 4).” Iacoboni et al. reported that the additional and different activation in Intention conditions would indicate mirror neurons specifically coding the intention of the agent. Relying on the signal level to support the intention coding hypothesis, Iacoboni et al. offered the following explanation: “In addition to the classically described mirror neurons that fire during the execution and observation of the same motor act (e.g., observed and executed grasping), there are [logical] neurons that are visually triggered by a given motor act (e.g., grasping observation), but discharge during the execution not of the same motor act, but of another act, functionally related to the observed act (e.g., bringing to the mouth) [1].
It appears, according to Iacoboni et al., that both classical and logical mirror neurons were triggered by the observation of a grasping motor act. However, the classical mirror neurons did respond as if they were triggered by the observation of the grasping motor act, but the logical mirror neurons responded as if they were triggered by observation of the “bringing to the mouth” motor act. This unusual pattern of mirror neuron response, which led to the additional increase of mirror neuron activity in the Intention drinking condition, could not be explained by Iacoboni et al. using Action Recognition Theory alone. Therefore, Iacoboni et al. suggest that there was another factor for the reported additional increase of mirror neuron activity (overlooking the difference between “grasping actions” A1 and A2), referred to as “logically related” mirror neurons. That other cause is understanding intention behind an action.
This proposed explanation was based on the assumption that the participants did not observe the subsequent “bringing to the mouth” motor act, but only the “grasping” motor act, and therefore the additional mirror neuron activity, which could not be derived from the observation of the “bringing to the mouth” motor act, became the basis for Iacoboni et al.’s conclusion: “The present findings not only allow one to attribute a functional role to these ‘logically related’ mirror neurons, but also suggest that they may be part of a chain of neurons coding the intentions of other people’s actions” [1]. Iacoboni et al. state that “neurons of this type have indeed been previously reported in F5 and referred to as ‘logically related’ neurons”[1]. They cite di Pellegrino to support his intention coding hypothesis: “Neurons of the rostral part of inferior premotor cortex of the monkey discharge during goal-directed hand movements such as grasping, holding, and tearing” [3]. Thus, di Pellegrino concluded that “premotor neurons can retrieve movements not only on the basis of stimulus characteristics, as previously described, but also on the basis of the meaning of the observed actions” [3]. However, in di Pellegrino’s experiments from 1992, the actions were surrounded by no context.
In contrast, Iacoboni et al. pointed specifically to the role of context as the cause for the activation of the logically related mirror neurons, and the understanding of intention: “Because “drinking” and “cleaning” contexts determined different activations in the Intention condition, it appears that there are sets of neurons in human inferior frontal cortex that specifically code the “why” of the action.” [1].
This raises the obvious question of how the “meaning of an action” is understood by monkeys if there is no context around the action. It also raises another concern of what exactly triggers the activation of the logical mirror neurons: the understanding of intention in context, as suggested by Iacoboni et al. (2005); the “meaning of an action,” without context, as suggested by di Pellegrino (1992); or the understanding of “intention without context,” as suggested by Fogassi et al. (2005), in their analysis of grasping actions: “Some of these ‘action-constrained’ motor neurons had mirror properties and selectively discharged during the observation of motor acts when these were embedded in a given action (e.g., grasping-for-eating but not grasping-for-placing)…. This specificity allowed the observer not only to recognize the observed motor act, but also to code what will be the next motor act of the not-yet-observed action: In other words to understand the intentions of the action’s agent” [7].
According to Fogassi et al., the observer (monkey) does not need context to understand the intentions of the agent. Fogassi et al. certainly seem to be at odds with Iacoboni et al.’s position that intention can be understood in Action without context: “Here the context cued the intention behind the action. Thus, the Intention condition contained information that allowed the understanding of intention, whereas the Action and Context conditions did not” [1].
Compounding the uncertainty, Iacoboni et al. offered multiple suggestions as to what the logically related mirror neurons are coding: intention, other potential motor acts, actions, or understanding of intention.
Q1. “In other words, in the Intention condition, there is activation of classical mirror neurons, plus activation of another set of neurons coding other potential actions sequentially related to the observed one.”
Q2. “The present findings not only allow one to attribute a functional role to these “logically related” mirror neurons, but also suggest that they may be part of a chain of neurons coding the intentions of other people’s actions.”
Q3. “The present findings strongly suggest that coding the intention associated with the actions of others is based on the activation of a neuronal chain formed by mirror neurons coding the observed motor act and by “logically related” mirror neurons coding the motor acts that are most likely to follow the observed one, in a given context.”
Q4. “The aim of the present study is to investigate the neural basis of intention understanding in this sense and, more specifically, the role played by the human mirror neuron system in this type of intention understanding.
Q5 “Because “drinking” and “cleaning” contexts determined different activations in the Intention condition, it appears that there are sets of neurons in human inferior frontal cortex that specifically code the “why” of the action and respond differently to different intentions.”
Iacoboni et al. suggested that mirror neurons code the “why” of an action, or “what comes next.” That requires understanding language concepts such as “the hand grasps the cup in order to drink,” which implies a mental thought process. Clearly, the answer to this question is not likely to come from monkeys, as Iacoboni et al. implied: “Experiments in monkeys demonstrated that frontal and parietal mirror neurons code the ‘what’ of the observed action (e.g., ‘the hand grasps the cup’)”[1]. Iacoboni et al. did not address, however, the issue of whether these neurons, or a subset of them, also code the “why” of an action (e.g., “the hand grasps the cup in order to drink”).
The varying propositions on how logically related mirror neurons are triggered and what they are coding causes uncertainty about how the findings should be interpreted. Lehrer (2006) echoes this concern: “A third concern is that it’s a leap to go from ‘mirror neurons represent actions’ to ‘mirror neurons represent meaning.’ I’m not sure this is the case” [8].
In addition, Iacoboni et al. straightforwardly pointed out that the applied logic would be plausible if premises for the argument are stipulated. Specifically, drinking and cleaning Contexts each individually produced the same mirror neuron activity, and the grasping actions in the intention conditions are the same: “This logic would hold only if there is no differential signal increase in the ‘drinking’ and ‘cleaning’ Context conditions, when no action is displayed…. Because in the Intention clips the same action was shown in two contexts (‘drinking’ and ‘cleaning’), one can test the intention-coding hypothesis by analyzing the signal increase during observation of the Intention clips.” [1].
But, in fact, Iacoboni et al. did not stipulate the premises for the argument: the grasping actions in the intention conditions were not identical, and thus Iacoboni et al.’s interpretation of the findings is called into question. “Our findings clearly show that coding intention activates a specific set of inferior frontal cortex neurons and that this activation cannot be attributed either to the grasping action (identical in both “drinking” and “cleaning” Intention clips) or to the surrounding objects, given that these objects produced identical signal increase in the “drinking” and “cleaning” Context clips, when no action was displayed.” [1].
As we proposed earlier, a complete grasping action consists of three phases, in which the last is a hand actually grasping an object. It can be suggested that the subsequent action, bringing to the mouth, of the complete drinking action may also be thought of as consisting of three phases. The first would be the object(s) in the hand, followed by the hand-arm movement that brings the object(s) to the mouth, and lastly, the parting of the lips for the object. Action A2, observed by the participants in the Intention drinking clip, in our view represents the last phase of the “grasping action” as well as the first phase of the “bringing to the mouth” action. Consider an analogous pattern within the human body. An elbow, for example, is simultaneously part of two adjacent parts of the body, the upper arm and the forearm.
Furthering the analogy of co-joined properties, let us focus again on “the bringing to the mouth” action in food-related situations in general. If we observe someone holding a cup of milk or tea at the lips, this can be viewed as the last phase of the bringing to the mouth action, as well as the first phase of the drinking action, and represent both.
Our position of the simultaneous recognition of two actions is corroborated by the experiments of Fogassi et al. (2005), who reported that neurons fired even “before the monkey observed the human model starting the second motor act (bringing the object to the mouth or placing it in a cup)” [7]. A similar pattern of firing of mirror neurons was recorded in Iacoboni et al.’s experiments, in the Intention drinking case leading to the higher activity in mirror neuron areas in the inferior frontal cortex, as the result of the visual recognition of both actions at the same time during the observation of a single clip. This could logically explain why mirror neurons behave as if they were triggered by the “bringing to the mouth” action. As Iacoboni et al. stated: “In addition to the classically described mirror neurons that fire during the execution and observation of the same motor act (e.g., observed and executed grasping), there are [logical] neurons that are visually triggered by a given motor act (e.g., grasping observation), but discharge during the execution not of the same motor act, but of another act, functionally related to the observed act (e.g., bringing to the mouth) [1].
Our analysis suggests that in the Intention drinking condition, there is possibly an activation of classical mirror neurons, which results from the recognition of the grasping action, plus activation of “logically related” neurons, which results from the recognition of the bringing to the mouth action.
Together, these two actions form two consecutive and adjoining links within the drinking action. Both recognized actions contribute to the double-digit increase of mirror neuron responses, as seen in Figure 4. Our interpretation is in exact correlation with the basic principle of functioning mirror neurons. As Iacoboni et al. (2005) stated: “A neuron discharging during the execution of grasping also fires during observation of grasping done by another individual. The characteristic property of most mirror neurons is the congruence between their visual and motor properties [1]. In all fairness, Iacoboni et al. admitted that “this property cannot account for the present findings, specifically, the differences in response observed between the drinking and cleaning Intention clips”[1].
Our interpretation of the findings is supported by the results of a survey in which twenty-three randomly selected subjects participated. They viewed a scene of a person’s hand is either “grasping” a teacup or an open water bottle without any context, similarly to the “grasping action” A2 in the Intention drinking clip. These subjects were then asked to describe what they had observed. Twenty of them used the word drinking to describe what they saw, and three described it as holding.
In reference to the Intention cleaning condition, the “unseen” following action is not even suggested by the researchers. As was pointed out earlier, Figures 3 and 4 do not depict data (the additional mirror neuron activity) in the Intention cleaning condition to support the notion that mirror neurons responded to the unseen following action, or that the following action was understood. Comparing both intention conditions, we can find that in one condition, Intention drinking, the mirror neurons demonstrated significant additional activity, and, thus, according to the interpretation of Iacoboni et al., the unseen following action was understood. In the other condition, Intention cleaning, the mirror neurons demonstrated lack of additional activity, suggesting that the unseen following action was not understood.
Two cases with two different patterns of response raise the question of how the understanding of the intention in Intention cleaning is derived, if there is no additional mirror activity exhibited to justify the position that the following action understood,.
And if the following action would be understood, it still will not explain why someone did this action. Knowing what comes after a first action(s) does not explain necessarily why exactly you do this action. The lack of the activity in Intention cleaning condition may imply that mirror neurons do not code intention and the high elevation of the signal in the Intention drinking condition may be attributed to a different cause. Our proposal offers a further explanation to why the recognition of the following action takes place in the Intention drinking condition, and not in the Intention cleaning condition.
The ABCDE Perceptual Model of “Grasping” Action
Bertenthal (2005) suggested that “when we observe a hand grasping an apple, the analyzed elements would be the hand, the apple, and the movement of the hand toward the apple” [6]. We propose that when we observe a hand grasping an apple (phases 1 through 3), the observed objects are the hand, the apple, the location of the apple, and the location of the hand within the three dimensional peripersonal space—that is, the space between object A (the performer) and object B (the apple). According to Gallese (2000), peripersonal space “is by definition a motor space, its outer limits being defined by the working space of different body effectors such as the head or the arms [and eyes]” [9].
In reality, the content of the observed scene is more complex than has been described so far. It involves all the elements of the action, including the performer, the hand, the arm, the apple, and also a food tray or table, the performer’s fingers (grip), and the performer’s mouth, all of which are objects with particular geometric characteristics within the three-dimensional peripersonal space. The location of the hand (object C) changes from the performer (A) toward the apple (B).
Gallese (1999) proposed that movement is coded in abstract terms. “What is coded,” he said, “is not simply a parameter such as force or movement direction, but rather the relationship, in motor terms, between the agent [A] and the object of the action [C]”[5]. In the case of grasping actions, we have the hand, C, whose location changes from point A to point B: the distance between A and C increases, and the distance between C and B decreases.
So far, three specific components involved in the action have been defined: the performer, A; the apple, B; and the hand, C. The location of these vary between A and B, within the motor space. Let us label motor (or peripersonal) space as E, and the direction of movement (toward or away from the body), as D, which can vary within E. All the components of the elemental ABCDE model are spatial. The locations of A, B, and C are encoded in the spatial frame E.
We propose that if mirror neurons are indeed to respond to “grasping actions,” the target location for the hand and eye movement for the grasping or the adjacent action (bringing to the mouth) must be part of the scene within the peripersonal space during the observation. (Observation is a vital component of the execution of the action, which utilizes one motor effector, eyes, while execution requires both the hand and the eyes).
Fogassi et al.’s (2005) experiments corroborate my position. Fogassi et al. reported that “mirror neurons of the inferior parietal lobe (IPL) respond differently to an observed grasping action… (e.g., for eating or for placing) [different target locations]”[7]. In both of Fogassi et al.’s cases, a hand is moving in different directions. In the first case, the target location is the mouth; in the second case, it is the cup. In reality, they both are “grasp-to-place” actions (in the mouth or in the cup). In both cases, the activation of mirror neurons is recorded corresponding to the following action. According to Fogassi et al.: “The activity of 41 mirror neurons was recorded in the parietal lobe (IPL) of two rhesus macaques. Of a total of 41 mirror neurons, 15 fired vigorously when the monkeys observed the “grasp-to-eat” motion, but registered no activity when exposed to the “grasp-to-place” condition. For four other mirror neurons, the reverse held true: they activated in response to the experimenter eventually placing the apple in the cup, but not to eating it” [7].
The results of these experiments support our position that the observer simply recognizes both actions: grasping and bringing to the mouth in the “grasp-to-eat” case; and grasping and placing an object in the cup in the “grasp-to-place” scenario. In each case, the target location for the following hand and eye movement after the grasping is present and plainly seen.
With respect to the Intention drinking condition, which yielded the higher mirror neuron response (see Fig. 4), the subsequent action is “bringing to the mouth,” a portion of which is observed by the participants. The target location for the “bringing to the mouth” action can be recognized by seeing the performer’s hand on food. (No other target location is offered in the experimental scene, as in Fogassi et al.’s second experiment, the “grasp-to-place” action.)
The observation of a portion of the action (bringing to the mouth) can most likely internally generate a visual representation of this action in the observer’s visual cortex. Our speculation is corroborated by Umiltà’s (2001) observation that “the [visual] motor representation of an action performed by others can be internally generated in the observer’s [visual and] premotor cortex, even when a visual [scene] description of the action is lacking [limited]” [10].
The concept of automatic recognition of the following action in the food-related action is further corroborated by DeVries, Visser, and Prechtl (1984), who demonstrated that the “hand to mouth” schema is an inborn pattern: “Ultrasonic scanning of fetuses,” they wrote, “shows that the movement of the hand to the mouth occurs between fifty and one hundred times an hour from 12–15 weeks gestational age” [11]. Gallager (2005) proposed that hand-to-mouth movement, analogous to the “bringing to the mouth” action, may be an aspect of an early, centrally organized coordination. Thus, suggesting that hand to mouth action is not intentional [12].
With respect to Intention cleaning, the target location for the following action is not present in the clip, and hence no additional mirror neuron activation is exhibited. The same logic applies to the Action, A1. This can explain why both conditions produced a similar level of mirror neuron activity (see Fig. 4). In both cases, the target location for hand and eye movement for the following action are not seen and most likely will be different than in food-related activity, away from the body, outside the peripersonal space.
Our proposition is that the target location for the next hand and eye movement, be it grasping food (as in Fogassi et al.’s experiments and Intention drinking) or grasping inedibles (as in Intention cleaning and Action, A1), is an essential component for automatic recognition and for mirror neuron activation. This is in agreement with Iacoboni et al.’s statement that mirror neuron additional activity occurs in the area where hand actions are represented: “Actions embedded in contexts, compared with the other two conditions [Intention drinking only], yielded a significant signal increase in the posterior part of the inferior frontal gyrus and the adjacent sector of the ventral premotor cortex where hand actions are represented” [1]. Our position is corroborated by Umiltà’s findings that “mirror neuron activation could be at the basis of action recognition” [10]. To validate our argument the level of neural activity in Action, (full cup), A2, without a context, should be measured and compared with level of action A1, as shown in Iacoboni 2005, Fig.4. The different level of the respond in may suggest that the recognition of both actions (grasping and bringing to the mouth) may takes the place during the observation of A2.
Of particular importance for Iacoboni et al. was to measure the differences in mirror neuron response between participants receiving explicit versus implicit instructions observing intention clips. The researchers stated that the lack of such top-down influences on the functioning of the mirror neuron system suggests that automatic processing is taking place (see Fig. 5). There was no differential activity measured when participants received explicit or implicit instructions. “Critically,” wrote Iacoboni et al., “the right inferior frontal cortex—the grasping mirror neuron area…, showed no differences between participants receiving Explicit instructions and those receiving Implicit instructions”[1].
In our interpretation, the results of the experiments showed no differences in mirror neuron response between Intention cleaning and Action (Fig. 4). Those results did not change as a function of explicit instructions when intention was clearly understood by half of the participants. The same, steady level of mirror neuron activity is exhibited in Intention cleaning and Action, and in both the implicit and explicit versions of Intention cleaning. Therefore, there is no evidence or indication to suggest that activity of the mirror neurons was modulated by top-down processes. As Iacoboni et al. note, “top-down influences are unlikely to modulate the activity of mirror neuron areas” [1]. Thus, they have no evidence to draw a conclusion that mirror neurons code for intention. Our interpretation of the results eliminates the emerging complications in Iacoboni et al.’s findings and establishes a single principle of functioning for mirror neurons for each and every link of the chain representing a complete action.
Mirror neuron systems apparently constitute action-recognition mechanisms rather than intention-coding mechanisms. Perhaps, there are different levels of understanding other people’s intentions, but this issue remains a subject for further investigation.
One final observation: In our understanding, mirror neurons form a perceptual-motor processing system that code the positions and configurations of body parts in relation to each other and to external objects, in peripersonal three-dimensional space, in the visual, auditory, and tactile modes, by mapping their motor counterparts. This motor processing system facilitates the coordination of different motor repertories (i.e., body parts)—particularly, eyes, hands, mouths, and ears.
Conclusion
Our analysis of the experiments of Iacoboni et al. (2005) demonstrates that it has not been proven that mirror neuron systems code intentions. The researchers’ findings that “observing grasping actions embedded in contexts yielded greater activity in mirror neuron areas in the inferior frontal cortex than did grasping actions in the absence of contexts or while observing contexts only” are not supported by the results of the experiment in the Intention cleaning condition. No sufficient additional increase is recorded to justify the finding of the understanding of intention. It appears that these findings are supported by the results of the experiment in the Intention drinking condition. However, the high response in the Intention drinking condition is due to the simultaneous recognition of two actions (grasping and bringing to the mouth) by viewing the Intention drinking clip. The lack of an additional response in the Intention cleaning condition is due to the recognition of the grasping action only, by viewing the Intention cleaning clip.
The various levels of responses in Action conditions can be attributed to the different types of grasping actions in the intention clips being observed by the participants. Those results are derived from the experimental model, but the arguments in support of the hypothesis are derived from the postulated model (identical grasping actions). This overlooked inconsistency led to an improper reading and interpretation of the data, thus compromising the validity of the findings.
The notion that context plays a significant role in support of the hypothesis is not proven. Furthermore, the role of context in the Intention drinking condition versus Action A2 (tea cup) could not be determined, because the level of the mirror neuron activity in A2 was not measured, and both conditions can be considered part of other plausible scenario. The promised comparison between Intention drinking (A1) and Action A1 was never done. Therefore, the findings that “the human mirror neuron system does not simply provide an action recognition mechanism, but also constitutes a neural system for coding the intentions of others” based upon the differential level increase in both intention conditions are not proven.
Our alternative interpretation can help to clarify our understanding of the mirror neuron functioning. Under our interpretation, the workings of the mirror neuron system are simplified under a single concise principle which supports action recognition theory.
References
1. Iacoboni Marco, Istvan Molnar-Szakacs, Vittorio Gallese, Giovanni Buccino,
John C.Mazziotta, Giacomo Rizzolatti. (2005). “Grasping the Intentions of
Others with One's Own Mirror Neuron System”. PLoS Biol. 2005 Mar; 3(3):e79.
2. Harmon-Jones Eddie, Piotr Winkielman - 2007 - Family & Relationships –
Social Neuroscience: Integrating Biological and Psychological .447-448 pages
3. Di Pellegrino G, Fadiga L, Fogassi L, Gallese V, Rizzolatti G (1992)
Understanding motor events: A neurophysiological study. Exp Brain91:176180.
4. Gallese V, Fadiga L, Fogassi L, Rizzolatti G.Brain. 1996 Apr;119 ( Pt 2):593-609 Action recognition in the premotor cortex, Italy
5. Gallese, Vittorio. (1999). “From Grasping to Language: Mirror Neurons and
the Origin of Social Communication.”Journal of Consciousness Studies,pp.1-10.
6. Bertenthal, B.I ., Motor Priming and Action Understanding. The logical
intersection of the previous two research themes concerns how the perception
and execution of actions
www.ccp.uchicago.edu/~bbertent/action_understanding.htm - 13k – (2005 -
data was taken from Internet)
7. Fogassi Leonardo Pier Francesco Ferrari,Benno Gesierich,Stefano Rozzi,
to Intention Understanding , Science 29 April 2005: Vol. 308. no. 5722,
pp.662-667
8. Lehrer, Jonah , “Small Gray Matters “
www.smallgraymatters.com/2006/07/25/mirror-neurons-arent-really-all-that-bad/
|
9. Gallese, Vittorio. (2000). “The Inner Sense of Action Agency and Motor
Representations.” Journal of Consciousness Studies, 7, no. 10, pp. 23–40.
10. Umiltà,M A, E Kohler, V Gallese, L Fogassi, L Fadiga, C Keysers, G
Neuron: 32, 91-101.
11 DeVries, Visser, and Prechtl 1984. Fetal motility in the first half of pregnancy.
Continuity of neural functions from prenatal to postnatal life (pp.46-64). London: Spastics International Medical Publications.
12. Gallagher Shaun How the Body Shapes the Mind, Oxford: Clarendon. Press, 2005.
No comments:
Post a Comment