Where are mirror neurons?
Mirror neurons were first identified in the premotor cortex of monkeys in 1992, and since that time they have also been found in several other areas of the monkey brain, including the primary motor cortex, inferior parietal lobule, frontal cortex, and the area surrounding a sulcus called the intraparietal sulcus.
There is very little conclusive evidence that mirror neurons exist in the human brain, although there is evidence from neuroimaging studies that indicates there are neurons in the human brain that display patterns of activity similar to the mirror neurons identified in the monkey brain. There is, however, one study to date that directly recorded the activity of purported mirror neurons in the brains of human patients who were being prepared for neurosurgery. Although this study could only explore certain areas of the brain (which didn't include the regions most frequently associated with mirror neurons in monkeys), investigators found neurons in the supplementary motor area and temporal lobe that displayed properties of mirror neurons. This, combined with the neuroimaging data mentioned above, suggests that mirror neurons likely exist in the human brain as well as the monkey brain.
What are mirror neurons and what do they do?
In 1992, a group of researchers at the University of Parma in Italy were recording the activity of individual neurons in the brain of a macaque monkey. They were observing neurons in the premotor cortex---specifically a region of the premotor cortex called area F5. Previous research had found neurons in this area to be active during goal-directed hand movements (e.g. grasping, holding, etc.). The investigators at the University of Parma were attempting to further understand this type of neural activation when they observed something surprising.
They noticed that neurons in the F5 region of the monkey's brain were activated not only when the monkey moved its hands, but also when the monkey observed an experimenter using his or her hands (e.g. to pick up a food reward and place it in the testing area). Four years later, they named these neurons mirror neurons because they seemed to be active not only when monkeys performed a particular action, but also when they saw someone else perform a similar action.
As mentioned above, since their initial discovery mirror neurons have been found in various other regions of the monkey brain (as well as in the brains of other species like songbirds), and there is evidence to suggest that mirror neurons exist in human brains.
The discovery of mirror neurons has generated a type of excitement both in and outside of the scientific community that is not often seen in response to scientific findings. Some have interpreted the activity of mirror neurons as the basis for our ability to understand the actions of others---a deduction thought by some to be unjustified, and one that has led to other (perhaps even less justifiable) extrapolations. For example, some have claimed that mirror neurons provide the necessary neural machinery for empathy, complex social interactions, language---and even that they are responsible for the rapid cultural advancement of the human race that led to us becoming modern humans.
Based on these proposed roles for mirror neurons, researchers began to speculate that impaired functioning of mirror neurons may be the basis for certain psychiatric disorders. For example, some have argued that dysfunctional mirror neurons underlie autism spectrum disorders (ASD). This hypothesis, sometimes called the "broken mirror hypothesis," suggests that individuals with ASD have abnormalities in mirror neuron networks that cause them to have an impaired ability to experience empathy, difficulty understanding the actions of others, and deficits in various aspects of social interaction ranging from eye contact to language.
Indeed, in the few decades since their discovery, mirror neurons have been credited or blamed for a long list of things ranging from simple feats like helping us to enjoy watching sports to complex emotions like compassion to disorders like ASD and schizophrenia.
The problem with these claims, however, is that they are mostly unsubstantiated. The first caveat to speculation about mirror neurons is that the vast majority of concrete evidence we have to support the existence of mirror neurons comes from studies in monkeys (concrete evidence in this case refers to evidence obtained from monitoring the activity of individual neurons---something that is difficult to do in humans except in rare circumstances like the example cited above where mirror neurons were explored in patients preparing for neurosurgery). Therefore, at this point we cannot with confidence attribute behaviors to mirror neurons unless we have been able to record mirror neuron activity while monkeys have exhibited such behaviors. Thus, we do not yet have the evidence to consider complex human emotions and behaviors (which can only be roughly approximated in studies of non-human primates) as attributable to mirror neurons.
Similarly, the evidence to point to dysfunctional mirror neurons as the main causal factor in human psychiatric disorders is lacking. Let's take the hypothesis that dysfunctional mirror neurons contribute to ASD as an example. This hypothesis was initially supported by two highly-cited studies from the early 2000s. One was a neuroimaging study that found reduced activity in autistic patients in a part of the brain thought to be heavily populated with mirror neurons. The other used electroencephalography to measure electrical activity believed to be indicative of mirror neurons; again, individuals with ASD appeared to display abnormalities in this activity.
Each of these studies, however, failed to replicate multiple times. Additionally, critics of the "broken mirror hypothesis" have been quick to point out that there is not good evidence that individuals with ASD even have deficits in understanding the intentions of others. Thus, the "broken mirror hypothesis" has been found to be wanting, and other hypotheses that attribute psychiatric abnormalities to mirror neurons are similarly in need of more support to make them tenable.
Even when it comes to just the basics of mirror neuron function, we are still searching for answers. For example, some researchers argue that the evidence that mirror neurons are involved with something as abstract as understanding actions is inadequate. According to this perspective, even if mirror neurons may be involved with functions like recognizing basic movements, selecting movements to make, etc., the evidence isn't conclusive to suggest mirror neurons are involved with a type of higher-level cognition like understanding the behavior of others. This is a critically important point as the idea that mirror neurons help us to understand others' actions is essential to the interpretation of mirror neurons as being involved in behaviors like empathy and social interaction---and indeed is the basis for much of the enthusiasm about mirror neurons in general.
Thus, while mirror neurons have been lauded for their ability to explain a variety of uniquely-human behaviors and accomplishments, it seems that we may have jumped the gun a bit on our interpretation of the activity of these cells. Much more research still needs to be done before we can say with confidence what mirror neuron activation really means in terms of behavior---and indeed before we can be sure that mirror neurons are as prevalent in the human brain as they are in the monkey brain. As with all scientific discoveries, it is best to be conservative in our interpretations until they are the only logical ones to make based on the data.
References (in addition to linked text above):
Hickok G. Eight problems for the mirror neuron theory of action understanding in monkeys and humans. J Cogn Neurosci. 2009 Jul;21(7):1229-43. doi: 10.1162/jocn.2009.21189.
Kilner JM, Lemon RN. What we know currently about mirror neurons. Curr Biol. 2013 Dec 2;23(23):R1057-62. doi: 10.1016/j.cub.2013.10.051.