Studies of friendship seem to implicate more complex factors. For example, one function fr

iend-ship seems to fulfill is 【C1】______ it supports the image we have【C2】______ ourselves, and【C3】______ the value of the attitudes we hold. Certainly we appear to【C4】______ ourselves onto our friends; several studies have shown that we judge them to be more like us than【C5】______. This suggests that we ought to choose friends who are similar to us【C6】______ than those who would be【C7】______ ("opposites attract"), a prediction which is supported by【C8】______ evidence, at least so【C9】______ as attitudes and beliefs are【C10】______. In one experiment, some developing friendships were【C11】______ amongst first-year students living in the same hostel. It was found that similarity of【C12】______(towards politics, religion and aesthetics) was a good predictor of【C13】______ friendship would be established by the end of four months, though it had less to do with initial【C14】______—not surprisingly, since attitudes may not be obvious on first inspection.

The difficulty of linking friendship【C15】______ similarity of personality probably【C16】______ the complexity of our personalities; we have many【C17】______ and therefore require a disparate group of friends to support us. This of course can explain why we may have two close friends who have little in【C18】______, and indeed dislike each other. By【C19】______ large, though, it looks as though we would do well to choose friends who【C20】______ us.

【C1】

A．that

B．which

C．what

D．if