How Humans Mentalize And See Through Others Minds Psychology Essay
Friedman and Petrashek (2009) found that in undertaking an ignorance prediction task, children did not follow the rule “ignorance means getting it wrong.” This experiment tested the work of Friedman and Petrashek. It involved using different amounts of information in each task to establish if what children are told affects how they see things from another’s point of view. Three conditions were used in this experiment with the amount of information about the character and what they do in the story differing in each condition. Each child undertook one condition of the ignorance prediction task, where they were told a story about a character who wanted to find an animal that was hiding under one of two boxes (the approach version) and then they were told another story about a character who did not want to find the animal and wanted to select the empty box (the avoid version). After being told the stories, the children were asked to point to one of the two boxes to indicate which one out of the two they thought the character would select to either find or avoid the animal. There was an equal chance that the children would select either box. Binomial tests were used to analyse the results in each of the conditions. The results indicated that the provision of information does have an effect and children do follow the “ignorance means getting it wrong” rule when they do not have sufficient information to predict success for the character. These results contrast with the findings of Friedman and Petrashek (2009). However, when there is enough information provided about the character, children were more likely to predict success.
As human beings, it is important that we learn the social skill of being able to understand the minds of others. The ability to do this is called mentalizing. This is when we, as humans, use mental representations of our and other people’s emotions, which allows us to see and interpret human behaviour in terms of deliberate mental states, such as feelings, desires, beliefs, goals etc (Fonagy, P., Gergely, G., Jurist, E.L and Target, M, 2002). We need mentalizing to be able to form beneficial attachments and have good relationships. According to Fonagy et al. (2002), securely-attached individuals are more likely to have had a mentalizing primary caregiver, which results in the child having a stronger capacity to represent the states of their own and other people’s minds. Furthermore, individuals with autism and other psychological disorders such as schizophrenia, can have deficits in the ability to mentalize. For these problems there are now mentalizing based therapies (clinical interventions) available for parents, infants, and young children (Slade, 2005).
The first studies to look at mentalizing and to establish at what age children could understand the minds of others were conducted by Wimmer and Perner (1983). This was known as the “unexpected transfer test”, and was used to establish whether children could hold a false belief. In the study, children were shown a story in which a character called John put a chocolate bar into a cupboard and then left the room. Whilst he was gone, his mother came into the room and moved the chocolate bar from the cupboard and put it into the fridge. The children are asked where John would look for the chocolate when he came back. Children have to ignore the fact that they know the chocolate has been moved and put themselves into John’s position: he still thinks it would be in the cupboard. Results showed that most of the children aged five years and above answered that John would look in the cupboard. Children aged four years and below said that he would look in the fridge, showing a very distinct developmental phase in acknowledging a false belief. One of the reasons why children below four have trouble attributing a false belief could be due to “selection processing” (Leslie, Friedman and German, 2004). In this theory, three to four year old children have a default tendency to attribute true beliefs to a character. Children must try and stop this tendency to attribute false beliefs and failure to do this causes three to four year olds’ to have difficulty in attributing false beliefs.
There are many different theories for how children answer a false belief task in the way that they do. The Simulation account (Goldman, 1989) illustrates that when looking at mental states, a person’s cognitive system mimics (simulates) the other person’s thoughts and feelings. When a character is portrayed as being ignorant, for example, the individual will mimic this ignorance and the individual will assume failure for the character. Furthermore, the Theory-Theory account (Gopnick & Wellman, 1992) states that there are nomological rules which help to explain events. These rules help to show which pieces of evidence are important and how they are interpreted. The evidence can also be changed and built upon if new information comes to light. Some of the cognitive development of children may be the result of gathering knowledge and generating new theories from old ones (Karmiloff-Smith and Inhelder, 1975).
Rules of conversation and judgement have an important impact in how children answer a false belief task. Schwartz (1994) believes that a participant will use the context of the conversation and the intent to resolve any ambiguities. If more information is given to participants about a character’s motives, for example, then they will be able to understand the situation and mind of the character better. This should hopefully lead to a successful outcome when they are asked questions about what they think a character will do. Furthermore, McCann & Higgins (1992) looked at a set of rules for communication. They state that the recipient of a communicator should take into account the characteristics of the communicator, determine the intent or purpose of what is being communicated, pay attention to the message and try to understand it. If children follow these rules of communication then the amount and type of information should have an impact on what they think a character will do. For example, if the intent or purpose of a story is not stated or explained very well, children will not have an accurate account of the character and will have difficulty in perceiving how the character is likely to think. Lastly, Ziegler, Mitchell and Currie (2005) found that the content and form of a narrative about a character, combine to give a strong cue to a child’s perspective, reinforcing the fact that the more information provided about a situation, the more likely a child is to predict the right outcome.
In contrast to the theories explained above, Ruffman (1996) argues that false belief may imply ignorance, but ignorance does not imply false belief. In his example, both Mike and Sally are looking for their dog. Mike is mistaken as to where the dog is, therefore he is ignorant of the location of his pet. Sally does not know where the dog is, but this does not imply that she is misguided, she may have no belief about the location of the dog and yet may luckily guess where it is. According to Ruffman (1996), 4 and 5 year old children do not grasp the asymmetry between ignorance and false belief. This is known as the “failure account” and when children are told about Sally (who does not know where the dog is), they should conclude that she has a false belief about where the dog is and so will fail when she looks for it. Ruffman (1996) tested this theory by showing children a scene with a dish of red and green sweets. They were shown a character who was told that a sweet was going to be placed into a box. The character went out of the room and the children saw that a green sweet was put into the box. When the children were asked which coloured sweet the character would think was in the box, they indicated that the character would get it wrong and think it was red, even though the character would equally likely choose red or green. The results are supported by Saxe (2005), who believes that children have a single concept for “not knowing” and “getting it wrong”.
However, children do not always follow the “ignorance means getting it wrong” rule. An experiment that demonstrates this was conducted by Hulme, Mitchell and Wood (2003), who told children that a character needs to find a teacher but they do not know what the teacher looks like. The children claimed that the character was ignorant of what the teacher looked like. However, when the children were shown pictures of faces and asked which best portrays the character’s belief about the teacher, they chose the picture of the actual teacher. Ignorance in this experiment does not necessarily mean that the character picks the wrong teacher.
Another experiment was conducted on this assumption by Friedman and Petrashek (2009), who used an ignorance prediction task in two experiments. The first experiment involved the participants being told a story about a girl named Sally who is looking for her dog. The participants were shown toy replicas of two boxes, with the dog hiding underneath one of the boxes. They were informed that Sally did not know which box her dog was under and they were asked where Sally would look for her dog. Even though there was an equally likely chance that Sally would look for her dog under both boxes, children were biased in selecting the box with the dog underneath. When adults were tested using the same method, Friedman and Petrashek (2009) found that they were unbiased and selected the box with the dog underneath it and the empty box equally. The second experiment had the same method as the first, except this time the participants were told to select a box when Sally (i) wants to give the dog a ball (the approach version) or (ii) when she does not want to (the avoid version). Results showed that children predicted success in both conditions and therefore, children do not follow the rule that “ignorance means getting it wrong.” This experiment supports the Theory-Theory account in that the children use nomological rules to help them understand the character and the situation and to make a decision based on the information given to them.
Friedman and Leslie (2004) also found the same results when they conducted a similar ignorance prediction task on children aged four to eight years old. The children were told a story about Sally who wanted to put her hat under a box. Before going to fetch her hat, Sally looked under the boxes and saw a dirty frog under one of them. After leaving to go and fetch her hat, the children were then told that the frog moves to a different box. They were asked the question “Which box will Sally put her hat under?” Most of the children selected the box without the frog underneath it, reinforcing the theory that children do not follow the “ignorance means getting it wrong rule.” This experiment also supports the Simulation and Theory-Theory accounts in that just because a character is ignorant children will not always predict that the character will fail in a task.
This experiment will examine the research by Friedman and Petrashek (2009). The aim is to check whether their theory is correct, by using three different versions of the story, with differing amounts of information provided to the participants to see if there is a “Chekhov’s Gun” effect. This effect is when a factor is introduced early in a story, but its importance does not become clear until later on. This will show whether the way that information is phrased and presented will have a consequence on how children see through another’s point of view. Will the factor help or hinder children and will it bias the “ignorance means getting it wrong rule?” The experiment will also be able to show which theories in how children mentalize, are the most accurate. The three conditions that will be used in the experiment are (i) a no checking and no information condition, (ii) a checking but no information condition and (iii) an information but no checking condition. Findings will also be compared with the results of adults from the Friedman and Petrashek (2009) experiment who have also completed this study previously, to see if there are any similarities or differences and if there are, why these occur. It is important to conduct research in this area because realising what another person is thinking and how it can make people more in tune with each other, is an important cognitive and social skill that we can learn much from.
The experimental hypothesis is that the different information provided will have an effect and the more information that is provided about the character, the more children will predict success for the character, as the character is proved not to be ignorant. The hypothesis will be tested by using ignorance prediction tests with three different conditions with different amounts of information. The children will be asked which box out of two the character would choose (with the chance being equally likely that the character could choose either box).
The design used in this experiment was a between subjects design. Each participant was randomly allocated to one of three story conditions, by the randomised order of their consent forms. The experiment had two independent variables, the first being the approach and the avoid versions of the task (two factors) and the other being the three stories. The dependent variable was which box out of a choice of two the participants selected in answer to the questions, with either box being correct.
The number of children that took part in the experiment was 60. The children were aged between 5 and 7 years old (the mean age being 6 years). The total number of female participants that took part was 34 and the total number of male participants that took part was 26. There were a total of 20 children in each of the three conditions. Children were recruited from Nottinghamshire schools.
Apparatus & Materials
The apparatus used in this experiment consisted of (i) the consent forms signed by the parents of the children (Appendix A), (ii) a laptop computer (with a touch screen cover) having a computer programme installed which contained images of the stories used in the experiment together with the scripts that the experimenter read to the children (Appendix B), (iii) instructions that were recited to each child before they undertook the experiment together with the answer sheets that the experimenter used to write down the children’s answers (Appendix C), (iv) a pen used by the experimenter to write down the answers, and (v) stickers given to the children as a reward for taking part.
Each child took part in one of the three conditions of the experiment but the way in which each condition was carried out was the same for each child. The experiment took place in a quiet room. The consent forms were randomised to prevent bias in the experiment and each child was selected according to the order in which their consent form came in the randomised pile, taking the next form from the top of the pile each time. For example, the first child was given condition 1, the second condition 2, the third condition 3, the fourth condition 1 etc until all 60 children had been tested. After recording the age and gender of the child, the experimenter read to the child the instructions for the experiment, which was the same for each child that took part. The instructions were “I am going to tell you two stories and I want you to listen to them very carefully because I am going to ask you some questions about them. You can touch what you think the answer is on my laptop screen. You can also stop the experiment if you want and you don’t have to tell me why you have stopped. Are you ready to begin?” When the child was ready a condition of the experiment was loaded up onto the laptop. The approach story was told first. The first condition was no checking and no information, where information is not given to the participant that the dog likes hiding under boxes. Billy does not check which box his dog is hiding under but the participant can see which box the dog is under. The second condition was checking but no information, where information is given to the participant that the dog likes to hide under boxes and Billy checks to see if his dog is under the boxes. The dog is not under any of the boxes at the time when Billy checks but appears after Billy has checked. The third condition was information but no checking, where information is given to the participant that the dog likes hiding under boxes. Billy does not check which box his dog is hiding under but the participant can see which box the dog is under. The conditions are the same for the avoid story except the character is now Toby and the animal is now a kitten.
Every child was told two stories in each one of the conditions, made up of an approach and an avoid version. In the approach version, the child is told that the character wants to find the animal under the box and in the avoid version, the character wants to find the box without the animal underneath it.
The experimenter read the script of the story to the child, whilst showing them the corresponding pictures of the story on the laptop. Examples of images in the story used can be seen below (Figure 1).
Figure 1: Examples of the images used in the experiment
There were differences between the approach scripts. In condition 1 (no checking and no information), the children were just told “Here is a story about a boy named Billy. Look what he has! It’s a …..bone. Billy also has a dog. Let’s see which box the dog is under. Is he under the green box? No. Is he under the blue box? Yes, there he is!” There is no extra information and no checking. In condition 2, the children are told that Billy checks to see if his dog is under a box “Here is a story about a boy named Billy. Look what he has! It’s a......bone. Billy wants to give the bone to his dog as a treat. Billy goes into the room to look for his dog. Look, there are two boxes here. What colour is this box? And what colour is this box? Billy checks if his dog is hiding under the boxes. Is the dog under the green one? No. Is the dog under the blue one? No. Now Billy goes to look outside. Look what happens whilst Billy is gone! His dog comes into the room and hides under the green box.” In the third condition, the children are told that Billy’s dog likes to hide under the boxes, but he does not check to see if his dog is there beforehand “Here is a story about a boy named Billy. Look what he has! It’s a ...... bone. Billy also has a dog. Billy’s dog likes hiding under these boxes. Let’s see which box the dog is under. Is he under the green box? No, is he under the blue box? Yes, there he is!” The differences in the avoid scripts were the same.
After the experimenter had read the script and shown the story on the laptop to the children, they were then asked the same questions in the same order every time. The approach questions were (i) a prediction question: “Which box will Billy go to with the bone?” (ii) a justification question: “Why will he look there?” (iii) a reality control question: “Which box is the dog under?” and lastly (iv) a know control question: “Does Billy know that they dog is there?” The avoid questions were (i) a prediction question: “Which box will Toby go to with the fish?” (ii) a justification question: “Why will he look there?” (iii) a reality question: “Which box is the kitten under?” and lastly (iv) a know control: “Does Toby know that the kitten is there?”
When asking the prediction question, an image with two boxes was present on the laptop screen (Figure 1). The child was asked to rest their hands on the keyboard and then to touch on the laptop screen which box they thought the character would go to. This was also the same for the reality control question. The child was asked to verbally explain their answers to the justification and the know control questions to the experimenter. The experimenter wrote down all the answers to the questions on an answer sheet, for each story (Appendix C). After the approach version story was told, the avoid version was undertaken in the same way (see Appendix B for the script and images used in the avoid version).When the child had completed both the approach and avoid versions of the story in their condition, they were thanked by the experimenter and allowed to choose a sticker as a reward for their participation.
The next consent form was picked and this child undertook a different condition. The same process in each of the conditions was used, until all 60 children had been tested.
After all the data was collected, the frequency data was summarised. The total number of children that selected the box with the animal underneath it on the approach task and selected the empty box on the avoidance task was found for each of the three conditions. Binomial tests were then used to analyse the data.
In condition 1 (no checking and no information) there was no significant difference in which box the children selected in the approach version (6 selected the empty box, 14 selected the box with the animal underneath) Z = 1.81, p > 0.05 and there was also no significant difference in which box the children selected in the avoid version (13 selected the empty box, 7 selected the box with the animal underneath) Z = 1.36, p > 0.05. In condition 2 (checking but no information) there was a significant difference in which box the children selected in the approach version (3 selected the empty box, 17 selected the box with the animal underneath) Z = 3.18, p < 0.05 and there was also a significant difference in which box the children selected in the avoid version (16 selected the empty box, 4 selected the box with the animal underneath) Z = 2.72, p < 0.05. In condition 3 (information but no checking) there was a significant difference in which box the children selected in the approach version (1 selected the empty box, 19 selected the box with the animal underneath) Z = 4.09, p < 0.05. However, there was no significant difference in which box the children selected in the avoid version (13 selected the empty box, 7 selected the box with the animal underneath) Z = 1.36, p > 0.05.
Figure 2: A bar chart to show the percentage of participants that predicted success in each of the conditions
The bar chart in figure 2 shows that there was no difference in which boxes were selected in both the approach and avoid story versions in condition 1. However, there were differences in which box the children selected in conditions 2 and 3 in the approach and avoid stories.
The results of the experiment show that how the stories were phrased in condition 2 (checking but no information) and condition 3 (no information but checking) had a significant effect in both the approach versions of the task, but only the avoidance version for condition 2. Participants predicted more success for the character in conditions 2 and 3, but not in condition 1 (no checking and no information).
From these results we can accept the experimental hypothesis. Certain information does have an effect in the responses of the children. Children in condition 1 were sensitive to the fact that the character is ignorant and so they do not necessarily predict success for the character. These results were the opposite of Freidman and Petrashek’s (2009) findings. In conditions 2 and 3, there is more information about the knowledge the character has and so, children are less likely to think that the character is ignorant and predict success.
This experiment has also been undertaken with adults (p.c, F. V. Ziegler, 2010). It was found that in the third condition (information but no checking) the adults did not predict overall success for the character, as they predicted approximately equal amounts of success and failure. In this experiment the results are slightly different in that in the approach version, mostly success is predicted by the children but in the avoidance version greater failure is predicted. This shows that children and adults use different information when looking at the minds of others. The results of the adults in this recent experiment also differ from the adults tested by Friedman and Petrashek (2009) who found that adults were unbiased about which box they selected in the ignorance predication test. This supports the theory that information does have an effect on how people see through the minds of others. More research could be undertaken to establish at exactly what age children begin to use the same cognitive processes as adults and what theories could explain the difference between the two and why this is so.
The fact that the children predicted failure for the character in the first condition (no checking and no information) but success in the other two conditions, could be due to a number of different reasons. The lack of information and context in the story, as explained by Schwartz (1994) could be a factor. The children were not told that the animal likes hiding under boxes and the character does not check whether his animal is under any of the boxes before he fetches the bone or fish. This may have caused the children to not fully understand the situation and so they thought that the character may not have understood the situation also and therefore predict less success for them, as in the Simulation account, where a child would mimic the thoughts and feelings of the character. Furthermore, using the Theory-Theory account as an explanation for the behaviour shows there was not enough information in the stories for the children to take into account and build upon and so, they could not predict for the character with the little information they had. Lastly, Ruffman (1990) may be correct with his “Failure account” i.e that when a character does not know where an object is, they will fail when looking for it.
Findings also rule out the “behavioural failure account” (Fabricius and Imbens-Bailey, 2000), which states that children cannot reason about the mental state belief and only seem to do so because they follow the rule “ignorance means your behaviour fails”. This experiment shows that children can grasp some belief about mental states and do not always follow the “ignorance means your behaviour fails” rule as success was predicted for the character in conditions 2 and 3 of the experiment.
The successful outcomes for the other two conditions where the character either checked the boxes before the animal hid underneath one or the participants were given the information that the animal likes to hide underneath the box, may be due to the fact that there is some information there that the child can use to help guide them as to how the character would act and if they would fail or succeed in selecting the correct box. For the checking condition, children may follow the rule by Saxe (2005) where children believe that the concept of seeing equals knowing and predict success. The character has checked before to make sure the animal is not underneath the box so this in the eyes of the children may mean that they deserve to find the animal as they have already looked before and it is not there, so when they look again they will find it.
The method used in this experiment could also be repeated but with children who usually fail false belief tests, such as children under four years old and children with autism for example, to see if having extra information could help guide them in attributing the correct belief to a character. This could be done to test whether theories like that of “selection processing” (Leslie, Friedman and German, 2004) are correct, as the results of this experiment have contradicted those found by Friedman and Petrashek (2009).
To conclude, children do follow the “ignorance means getting it wrong” rule when they do not have enough information to predict success for the character, which is the opposite of what Friedman and Petrashek (2009) found. However, when children are provided with more information, they will predict success for the character, showing that the “Chekhov’s Gun” theory does have an effect in how children understand ignorance and what they predict a character will do. These findings were different to the adults that were tested on this experiment, in that the adults predicted less success for the character in the information but no checking condition than the children. This could be due to many different reasons, for example adults and children use different theories when mentalizing or adults having a more sophisticated cognitive processing and understanding and the children that were tested not having developed this yet. How humans mentalize and see through the minds of others is a very complicated yet precise cognitive task. More research needs to be carried out in this area, as there are conflicting results among studies and a specific answer as to how children see through the minds of others should be found.