Scientists at Tel Aviv University claim to have created neuronal memories on a silicon chip. Live neurons were put on a silicon chip that had electrodes for reading electrical activity. Every time scientists put a nerve-stimulating chemical at the same spot, they saw the same pattern of electrical activity come out of the electrodes, then die down. After several repetitions, the pattern continued without further chemical stimulus. The researchers believed the neurons learned to anticipate the chemical and claimed that the neuron group had formed a memory.
But there are two things wrong with the analogy and the conclusion. First, there was no conditioned stimulus, the equivalent of Pavlov’s bell (he actually used a buzzer, but the idea of a bell has become fixed in folklore). Pavlov paired the bell and the food many times, then found that the dog would salivate to the bell alone. (Pavolv's Nobel Prize acceptance speech about this topic was scorned as "too mental," not scientific).
In the neural cell assembly scenario, the neurons had nothing to anticipate. There was no bell (and neurons can't hear anyway). They merely perseverated their previous activity. A plucked guitar string will continue to sound a tone for a while, but that does not demonstrate learning or memory, at least not in the cognitive sense of memory.
The second problem is with this study's conclusion. The authors assume that memory is a certain pattern of neural activity. But that definition plays on a semantic ambiguity. An alarm clock has memory, but that is a functional use of the term. If we mean cognitive memory, as humans have, then the alarm clock doesn't have it, and neither do the cells on a chip. A cognitive memory is a re-experience.
My memory of last night’s dinner includes lemon, risotto, and Syrah. It does not have any quality of a cell assembly,which is not an explanation adequate to the phenomenon. Pointing out a neural correlate to memory is helpful, but naming cell activity, literally, “a memory,” is thoughtless or malicious misdirection.
Overall then, the interpretation of this study is utterly confused. It has nothing to do with memory. Don't believe everything you read!
D.C. (2007). This is your brain on a chip. Science News, 171, (April 21), 253.