Hobson and colleagues' study is based on the assumption that in the waking we constantly think in a logical, purposeful and empirically relevant way, which is not the case. There are various degrees of thought control in the waking and consequently different degrees of rationality in the mind's productions. The conclusions of the Hobson and colleagues' study might have been very different if the authors had compared dream reports with similar products of the waking mind. For instance, spontaneous remembrances and anticipations share several features with dream reports and informal oral descriptions of an autobiographical episode have a similar sequential organization. Daydreaming includes bizarre elements, abrupt changes of topic, and sometimes a loss of reality testing. Dreaming is producing world simulations, in other words imagining. Like the products of waking imagination it is not devoid of unrealistic aspects and discontinuities. In order to understand why a dreamer imagined a certain event, we must take into account that the human mind is prone to use metaphors. Dreaming has to use metaphors because it cannot literally represent abstract ideas and long and intricate plots. It has to replace them by concrete and rather simple, short, and homogeneous events. Although I disagree with Hobson and colleagues' method, I am happy to see that the gap between their conception of dreaming and the views of cognitive psychologists is narrowing, now that they admit dreaming is not totally or essentially irrational.