Instead of throwing light on subject, dual concepts of feeling and sensation seem to confound matters even further. A more basic level needs to be broached, that of sense data (or sensa, as in this text).

Sense data are entities cyclically defined. Their existence depends upon being sensed by a sensor equipped with senses. Yet, they define senses to a large extent (imagine trying to define sense of vision without visuals). Ostensibly, they are entities, though subjective. Allegedly, they possess properties that we perceive in an external object (if it is there), as it appears to have them. In other words, though external object is perceived, what we really get in touch with directly, what we apprehend without mediation – are subjective sensa. What is (probably) perceived is merely inferred from sense data. In short, all our empirical knowledge rests upon our acquaintance with sensa. Every perception has as its basis pure experience. But same can be said about memory, imagination, dreams, hallucinations. Sensation, as opposed to these, is supposed to be error free, not subject to filtering or to interpretation, special, infallible, direct and immediate. It is an awareness of existence of entities: objects, ideas, impressions, perceptions, even other sensations. Russell and Moore said that sense data have all (and only) properties that they appear to have and can only be sensed by one subject. But these all are idealistic renditions of senses, sensations and sensa. In practice, it is notoriously difficult to reach a consensus regarding description of sense data or to base any meaningful (let alone useful) knowledge of physical world on them. There is a great variance in conception of sensa. Berkeley, ever incorrigible practical Briton, said that sense data exist only if and when sensed or perceived by us. Nay, their very existence IS their being perceived or sensed by us. Some sensa are public or part of lager assemblages of sensa. Their interaction with other sensa, parts of objects, or surfaces of objects may distort inventory of their properties. They may seem to lack properties that they do possess or to possess properties that can be discovered only upon close inspection (not immediately evident). Some sense data are intrinsically vague. What is a striped pajama? How many stripes does it contain? We do not know. It is sufficient to note (=to visually sense) that it has stripes all over. Some philosophers say that if a sense data can be sensed then they possibly exist. These sensa are called sensibilia (plural of sensibile). Even when not actually perceived or sensed, objects consist of sensibilia. This makes sense data hard to differentiate. They overlap and where one begins may be end of another. Nor is it possible to say if sensa are changeable because we do not really know WHAT they are (objects, substances, entities, qualities, events?).

Other philosophers suggested that sensing is an act directed at objects called sense data. Other hotly dispute this artificial separation. To see red is simply to see in a certain manner, that is: to see redly. This is adverbial school. It is close to contention that sense data are nothing but a linguistic convenience, a noun, which enables us to discuss appearances. For instance, "Gray" sense data is nothing but a mixture of red and sodium. Yet we use this convention (gray) for convenience and efficacy's sakes.

B. The Evidence

An important facet of emotions is that they can generate and direct behaviour. They can trigger complex chains of actions, not always beneficial to individual. Yerkes and Dodson observed that more complex a task is, more emotional arousal interferes with performance. In other words, emotions can motivate. If this were their only function, we might have determined that emotions are a sub-category of motivations.

Some cultures do not have a word for emotion. Others equate emotions with physical sensations, a-la James-Lange, who said that external stimuli cause bodily changes which result in emotions (or are interpreted as such by person affected). Cannon and Bard differed only in saying that both emotions and bodily responses were simultaneous. An even more far-fetched approach (Cognitive Theories) was that situations in our environment foster in us a GENERAL state of arousal. We receive clues from environment as to what we should call this general state. For instance, it was demonstrated that facial expressions can induce emotions, apart from any cognition.

A big part of problem is that there is no accurate way to verbally communicate emotions. People are either unaware of their feelings or try to falsify their magnitude (minimize or exaggerate them). Facial expressions seem to be both inborn and universal. Children born deaf and blind use them. They must be serving some adaptive survival strategy or function. Darwin said that emotions have an evolutionary history and can be traced across cultures as part of our biological heritage. Maybe so. But bodily vocabulary is not flexible enough to capture full range of emotional subtleties humans are capable of. Another nonverbal mode of communication is known as body language: way we move, distance we maintain from others (personal or private territory). It expresses emotions, though only very crass and raw ones.

And there is overt behaviour. It is determined by culture, upbringing, personal inclination, temperament and so on. For instance: women are more likely to express emotions than men when they encounter a person in distress. Both sexes, however, experience same level of physiological arousal in such an encounter. Men and women also label their emotions differently. What men call anger – women call hurt or sadness. Men are four times more likely than women to resort to violence. Women more often than not will internalize aggression and become depressed.

Introduction

This essay is divided in two parts. In first, we survey landscape of discourse regarding emotions in general and sensations in particular. This part will be familiar to any student of philosophy and can be skipped by same. The second part contains an attempt at producing an integrative overview of matter, whether successful or not is best left to reader to judge.

A. Survey

Words have power to express speaker's emotions and to evoke emotions (whether same or not remains disputed) in listener. Words, therefore, possess emotive meaning together with their descriptive meaning (the latter plays a cognitive role in forming beliefs and understanding).

Our moral judgements and responses deriving thereof have a strong emotional streak, an emotional aspect and an emotive element. Whether emotive part predominates as basis of appraisal is again debatable. Reason analyzes a situation and prescribes alternatives for action. But it is considered to be static, inert, not goal-oriented (one is almost tempted to say: non-teleological - see: "Legitimizing Final Causes"). The equally necessary dynamic, action-inducing component is thought, for some oblivious reason, to belong to emotional realm. Thus, language (=words) used to express moral judgement supposedly actually express speaker's emotions. Through aforementioned mechanism of emotive meaning, similar emotions are evoked in hearer and he is moved to action.

A distinction should be – and has been – drawn between regarding moral judgement as merely a report pertaining to subject's inner emotional world – and regarding it wholly as an emotive reaction. In first case, whole notion (really, phenomenon) of moral disagreement is rendered incomprehensible. How could one disagree with a report? In second case, moral judgement is reduced to status of an exclamation, a non-propositional expression of "emotive tension", a mental excretion. This absurd was nicknamed: "The Boo-Hoorah Theory".

There were those who maintained that whole issue was result of mislabeling. Emotions are really what we otherwise call attitudes, they claimed. We approve or disapprove of something, therefore, we "feel". Prescriptivist accounts displaced emotivist analyses. This instrumentalism did not prove more helpful than its purist predecessors.

Throughout this scholarly debate, philosophers did what they are best at: ignored reality. Moral judgements – every child knows – are not explosive or implosive events, with shattered and scattered emotions strewn all over battlefield. Logic is definitely involved and so are responses to already analyzed moral properties and circumstances. Moreover, emotions themselves are judged morally (as right or wrong). If a moral judgement were really an emotion, we would need to stipulate existence of an hyper-emotion to account for moral judgement of our emotions and, in all likelihood, will find ourselves infinitely regressing. If moral judgement is a report or an exclamation, how are we able to distinguish it from mere rhetoric? How are we able to intelligibly account for formation of moral standpoints by moral agents in response to an unprecedented moral challenge?

Moral realists criticize these largely superfluous and artificial dichotomies (reason versus feeling, belief versus desire, emotivism and noncognitivism versus realism).