The discursive dilemma

To my mind, one of the most intriguing issues in social philosophy is that groups as such can have different opinions than its individual members taken together have. It is not simply an ivory tower problem, for it can have practical consequences, for instance when a group takes a decision that is against the will of its members or when it takes a decision that doesn’t find enough support among its members, so that it is difficult to get it executed. I have discussed this problem already before in these blogs, but when I noticed that the last time I did is already six years ago (see here and here), I thought that it would be worthwhile to raise the matter again, but then from a somewhat different angle.
The problem has been famously discussed by Lewis A. Kornhauser and Lawrence G. Sager in their article “The one and the Many” in which they analysed the case of a three-member court that passes a verdict that deviates from what the individual judges think. But here I prefer to discuss an example treated by List and Pettit (2013, pp. 45-46), which is more general, because unlike judges, the participants are not limited by exogenous constraints like official procedures in their decisions. List and Pettit call this more general problem the “discursive dilemma”. Here it is, a little bit adapted by me:
  “[I]magine an expert panel that has to give advice on global warning. … The panel seeks to form judgments on the following propositions (and their negations):
- Global carbon dioxide emissions from fossil fuels are above 6500 million metric tons of carbon per annum (proposition ‘p’).
- If global carbon dioxide emissions are above this threshold, then the global temperature will increase by at least 1.5 degrees Celsius over the next three decades (proposition ‘if p then q’).
- The global temperature will increase by at least 1.5 degrees Celsius over the next three decades (proposition ‘q’).
The three propositions are complex factual propositions on which the experts may reasonably disagree. Suppose the experts’ judgments are shown in the table below, all individually consistent. … 

                        Emissions above         If p then temper-         Temperature
                        threshold?                   ature increase?            increase?
                        (p?)                             (if p then q?)               (q?) 

Individual 1                True                True                            True 

Individual 2                True                False                           False 

Individual 3                False               True                            False 

Majority                     True               True                           False

Given the judgments in this table, a majority of experts judges that emissions are above the relevant threshold (‘p’). Moreover, a majority judges that, if they are above this threshold, then the temperature will increase by 1.5 degrees Celsius (‘if p then q’). Nevertheless, a majority judges that there will be no temperature increase (not ‘q’).” List and Pettit conclude then that “a majority voting on interconnected propositions may lead to inconsistent group judgments even when individual judgments are fully consistent …” (p. 46).
I think that examples like this one illustrate that groups are not simply aggregates of individuals. Groups are not just collections of certain individuals but they are entities of their own and in a sense they are independent of the individuals that make up the group. Otherwise, we couldn’t explain, for instance, how a sports team can become champion, if the members that make up the team at the beginning of the season are not the same members that make up the team at the end of the season (or for a part). Just as we don’t get a new car, when its tyres are replaced, we don’t get a new team when one or more members are replaced. That a team becomes champion is the consequence of purposeful and intentional actions by the team members but as such these individual actions aren’t actions of the group. A team can play a match because its actions are constituted by the individual actions of its players, but the players of the team don’t need to be the same players all the time. Therefore, in the end, it’s not that the individual players win the cup but the team does. I think that something like this happens when a group takes a decision, as in the example above. The group members think individually and vote individually and this results in a group decision, but this individual voting is not the same as the group decision. It would be different if the group members would take a decision in a joint consultation in consensus.
The case discussed exemplifies that generally what groups do and what individuals do are different things and are on different levels. Groups are not simply individuals put together. This is an important conclusion. If one doesn’t take it into account, it can happen, for instance, that a group takes a decision that cannot be executed because it is against the will of its members, although all members had a say in it. 

Source
Christian List; Philip Pettit, Group Agency. The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press, 2013.