Decision-making signals in the primate parietal cortex

Lateral intraparietal area (LIP) neurons are thought to compute the decision of where to look. Specifically, their firing rate is thought to represent accumulated evidence for a decision by ramping up to a high level, the putative decision bound, before an eye movement to a given location. However, LIP neurons are also highly responsive to visual stimuli. Because all previous research put a visual stimulus (a saccadic target) in the response field (RF) location of a neuron during decision formation, it is unknown if LIP neurons can still show decision computation without this visual drive. We therefore recorded the spiking activity of LIP neurons in a conventional decision task where a monkey decides the direction of a noisy motion stimulus and indicates his decision with a saccade. On half the trials, the Choice Targets remained on for the whole trial, as is conventionally done. On the remaining trials, targets were flashed at the beginning of the trial and absent during motion-viewing. Furthermore, we recorded the activity of any neuron with an RF, instead of only neurons exhibiting persistent memory activity before an instructed saccade. This enabled us to also test the long-held assumption that only cells with persistent memory activity show decision signals. Our results show that 1) cells without persistent activity indeed show decision signals, 2) population response drops without RF stimulation (although individual neurons were affected in diverse ways), 3) distinct, repeating response “motifs” exist among cells, 4) a signal exists where neural response is lower for stronger motion strength stimuli, regardless of direction. These results prove that contrary to dogma, a neuron’s ability to show a long “memory” response is not related to an ability to accumulate evidence over time for a decision. Also, LIP firing rate cannot be interpreted as a pure decision variable because it simultaneously represents decision-irrelevant, visual stimuli. Finally, diverse, but repeating responses among cells suggest the existence of cell types in LIP. These results demonstrate that LIP acts as a bank of potentially useful signals, and raises the question of how they might be used for a decision.