Research into problem solving began in the late 1800’s with Thorndike (1898), and his work on cats in a puzzle box. In this experiment, a cat was placed in a box, and could only escape by revealing the appropriate device and action that would cause a trap door to open, for example, clawing at a lever. At first, the cat performed random activities until it accidentally uncovered its method of escape. However, after time and practice, “the cat tended to eliminate the unsuccessful responses and to escape more quickly from the puzzle box” (Eysenck, 1984).
Therefore, Thorndike proposed that problem solving arises due to trial and error, and speed of problem solving can be enhanced by practice. The Gestalt Psychologists put an alternative explanation forward in the early 1900’s. They too used animal studies to investigate problem solving, and argued that in order to solve a problem, it required us to reorganise elements of the problem situation. A famous experiment by Wolfgang Kohler demonstrated this. He studied apes kept in a cage, which with the use of a stick and a crate attempted to reach bananas hanging from the roof of the cage.
Kohler suggested that after a period of intense thinking, the apes then experienced a sudden insight that helped to solve the problem they had with reaching the bananas, by standing on top of the crate. These two accounts of problem solving are very different, and both have been heavily criticised. Thorndike’s ideas were criticised as it was argued that random processes cannot simply explain problem solving, and that there must be some method involved. Also, many did not accept the Gestaltist’s claims that insight is a regular occurrence in problem solving.
They were both also criticised for their failure to explain in depth, the processes actually involved in problem solving. This led to suggestions that past experience plays a huge role in problem solving in what is termed the ‘positive transfer effect’. This is where past experiences positively assist with present experiences and problem solving. However, a positive effect did not always occur, and a phenomenon known as ‘functional fixedness’ was also suggested, this being when prior knowledge interferes with, and disrupts later attempts to solve problems.
This was investigated by Duncker (1945). Participants were given objects including a candle and a box of nails, and were asked to attach the candle to a wall. Many tried methods such as trying to nail the candle directly to the wall, but very few thought of using the box as a candle holder and nailing this to the wall. Duncker explained this in terms of functional fixedness, suggesting that participants became “‘fixated’ on the box’s normal function of holding nails and could not re-conceptualise it in a manner that allowed them to solve the problem” (Eysenck & Keane, 1995).
By considering past experiences on problem solving, we are specifically looking at the speed and accuracy of problem solving. However, a criticism of such research is that it is difficult to speculate on the extent of the participants past experience. Newell & Simon (1972) went on to concentrate on the mental processes involved in problem solving, with their ‘General Problem Solver’ computer theory. This theory proposed that each problem is represented by a ‘problem space’, “the choices that the problem solver evaluates while solving the problem” (Reed, 2000).
These consist of the initial state, goal state and all intermediary states of the problem. The main method used for searching through the problem space is known as means-ends analysis, which involves “the problem solver generating goals and attempting to find operators capable of satisfying each goal. If a particular goal cannot be satisfied directly, then a sub-goal is created” (Eysenck, 1984). Subgoals exist between the initial state and goal state, and are located on the solution path. Subgoals are assumed to be helpful, as by knowing one exists, it is possible to avoid searching through the unwanted paths.
However, the subgoals are not always obvious, and require careful selection. The main idea of the theory is that problems are solved by completing a goal directed search through the problem spaces with help from planning and use of subgoals. However, the model was criticised for being unable to explain how experience is involved in human problem solving. It could not account for the fact that problem-solving strategies can change and become more efficient due to practice. It was also argued that the theory ignored the importance of forward planning in problem solving, and could not be applied to a very wide range of human problems.
Despite this, the main idea of how we solve problems by going from present to desired states appears to be of great importance in problem solving activities. There are many distinct differences present between the reasoning and problem solving research areas. It is obvious that there is a greater emphasis on the problem solving research, simply by the fact that the research into this area is much more in-depth. With the reasoning research, the main focus is on deductive inference, whilst almost ignoring the inductive reasoning.
However, this is not the case with the problem solving research, where all areas are considered to be of importance, and therefore it has been treated as a hole. Despite both reasoning and problem solving research both being areas of great interest back in history, it appears that the problem solving research has been brought on a lot more in recent years, making it a much more modern account. However, it may be that things are beginning to change, with psychologists altering their opinion that a distinction should be made between the two.
It was often thought that deductive reasoning involved logic, whereas problem solving does not. But when considering both the task requirements and the mental processes involved in problem solving, it appears that “logical reasoning tasks are very frequently tackled in an alarmingly illogical way” (Eysenck, 1984). This has led to the proposal that deductive reasoning is simply a special case of problem solving and that the two processes are combined, rather than the earlier beliefs that a distinction should be made between the two concepts.
Newell (1980) in particular, has fought for a unification of the two areas, arguing that the ‘problem space’ is present in all goal-directed behaviour, including both reasoning and problem solving. It now seems likely that in future references, reasoning and problem solving strategies will be explained as one, and although the necessary empirical evidence and theories are not firmly established as yet, they will be in years to come as reasoning becomes more accepted as a special case of problem solving by psychologists.