Antiphase or push-pull inhibition was first proposed by Hubel and Wiesel (1962), though they did not discuss its importance for contrast invariance. There is now ample evidence both from extracellular (Palmer and Davis, 1981, Heggelund, 1981) and intracellular (Anderson et al., 1999, Ferster, 1988, Borg-Graham et al., 1998, Hirsch et al., 1998) recordings that simple cells do, in fact, receive strong OFF inhibition in their ON subfields, and strong ON inhibition in their OFF subfields. This push-pull inhibition is stronger than relay-cell excitation, as required by Troyer et al. (1998)'s model. This was demonstrated in intracellular recordings by Hirsch et al. (1998), who showed that push-pull inhibition can completely suppress the response of simple cells to excitatory inputs. When a flashed spot of lightthat evoked a strong excitation from one subregion was moved slightly to encroach on a neighboring subregion of opposite polarity, the resulting push-pull inhibition overwhelmed the excitatory response, instead yielding hyperpolarization.
The overall dominance of cortical inhibition over excitation is also suggested by other experiments. Electrical stimulation of the LGN (Ferster, 1986, Douglas and Martin, 1991) or of the cortex (Chung and Ferster, 1998, Hirsch and Gilbert, 1991) evokes a brief excitation followed by a long lasting and often stronger inhibition. The same is often true of briefly flashed visual stimuli (Hirsch et al., 1998). Measurements of excitatory and inhibitory conductances evoked by drifting gratings show that the latter can be 2-5 times as large as the former (Anderson et al., 1999).
Push-pull inhibition is by far the dominant if not the sole form of inhibition received by simple cells. Any inhibition that is not in a push-pull arrangement, e.g. ON inhibition in an ON subfield, must be far weaker. To see this, note that a light spot flashed in an ON subregion evokes a strong depolarizing response. Therefore any inhibition evoked by the spot must be much weaker than the evoked excitation. Another important feature of the inhibition received by simple cells is its orientation tuning. Intracellular recordings show that this inhibition is maximal at the preferred orientation and falls off strongly away from the preferred. The orientation tuning width of the inhibition received by a cell appears to be nearly identical to that of the cell's excitatory inputs (Anderson et al., 1999, Ferster, 1986, Nelson et al., 1994, Douglas et al., 1991).