Thorndike’s Law of Spread of Effect principle states that reward strengthens not only the response to which it belongs, but also the responses adjacent (after or before) to the rewarded response. It gives rise to a gradient effect. The effect of reward is maximal for the rewarded response. Then its effect decreases for each step that a response is removed from the rewarded one. Because of temporal proximity, reward may even strengthen adjacent punished connections. If a response sequence, R1-R2-R3-R4-R5 brings reward S, S-R5 connection will obviously be strengthened. At the same time, the S would also strengthen the responses R1 R2, R3, R4 such that S-R4>S-R3> S-R2>S-R1.