286 Michael Spitzer emotions: Happiness, Tenderness, Sadness, Anger, and Fear. Another Darwinian tenet is that emotion is behavioural, involving a variety of ‘action tendencies’. Ac-cording to Juslin and Sloboda, ‘action tendency’ describes how an emotion ‘directs the person experiencing it towards one category of behaviours rather than another’ (Juslin and Sloboda 2001, p. 87). Thus a fearful person may respond to fear in a series of actions: closing her eyes; running away (fight); hiding; or confronting and attacking the fearsome object ( ght). These action tendencies are different from those associated with sadness (e. g. withdrawing in order to mourn or refect) or an-ger (e. g. attacking or threatening to attack). It is important to stress that action ten-dencies also bare upon cognition and physiology, since this will afford touchpoints with my arguments about musical thought and feeling. Thus, in their review of the adaptive functions of the discrete emotions, Izard and Ackerman show that Sadness slows the cognitive and motor systems enabling ‘a more careful look for the source of trouble and deeper refection on a disappointing performance’ (2004, p. 258). An-ger, by contrast, mobilizes and sustains ‘energy at high levels’ (p. 259), preparing the organism for a possible confrontation.I need to build a couple of metaphorical bridges, if this is to be relevant to music analysis. First, to relate emotional behaviour to the formal ‘behaviour’, as it were, of a musical work. That is, hearing the processes and gestures unfolded by the work as akin to those of a person. This involves two metaphorical transactions: hearing the music ‘as’ a persona; and hearing the music’s actions ‘as’ those of this persona.Critics, philosophers, and programme-note writers have long spoken of symphonies and sonatas dancing ‘happily’, singing ‘tenderly’ or drooping ‘sadly’. My approach simply contextualizes this tendency within a Darwinian intellectual paradigm. NB: I should point out the important distinction, voiced by Juslin et al, between emotion designated, and emotion inducted (or aroused). As in Peter Kivy’s (1989) famous ex-ample of a Saint Bernard’s ‘sad’ face, there cannot be any question of the dog really feeling sad (it may be, but we have no way of knowing that). Rather, the dog’s fea-tures are semiotically ‘expressive of’ sadness. In a similar way, Dido’s lament, from Purcell’s Dido and Aeneas, ‘designates’, or is ‘expressive of’, sadness, by utilizing and manipulating musical features associated with this emotion, such as slow tempo, minor mode, descending contour, and piano dynamics. It certainly doesn’t express the performer’s or composer’s possible sadness. Nor does it necessarily arouse sad-ness in the listener, who may, on the contrary, gain much joyful pleasure from listen-ing to the music. In short, I am limiting my discussion to music’s designated emo-tional behaviour. The second kind of metaphorical bridge is more speculative, since it bears on the evolutionary origins of emotions. In a nutshell, what relevance does a primitive ad-aptive emotion such as Fear, evolved millions of years ago, have for the aesthetic ex-perience of classical music today? I would like to here draw on the ideas of the evol-utionary psychologist, Keith Oatley. I propose a metaphorical relationship between ancient and modern emotions, on the basis of a surface/depth model. Theorists since Darwin have modeled the evolutionary development of the human brain on