Sentimentality may be described as: 

The attribution of a degree of significance or value to a phenomenon without sufficient evidence.

This definition is sufficiently general to capture all cases and provides a criterion for the determination of borderline disputes. Furthermore, though abstract it is not without important implications that are consistent with our intuitions, and revealing of aspects that are otherwise obscure.

For example, any appeal whatsoever to transcendent significance is classed as sentimental since the evidence of transcendence is weak at best and by rigorous standards quite non-existent. Thus the definition captures the palpable similarity between notoriously sentimental material such as romantic literature, on the one hand, and less obviously sentimental material such as religious texts on the other, a similarity that is otherwise hard to articulate since the immediate materials vary considerably in character and may even seem contradictory or at least very confused.

For example, religions in many instances forbid or restrict carnal affection, bringing them into conflict with the sensuous appeals of the romantic novelist, while at the same time employing the language of uncontrolled romantic love and loyalty to describe the relation between the human and divine. Consequently, while there is a strong affinity between romantic writing and religion in one area, this resemblance is cancelled out by the moral controls implied in another. Nevertheless, the intuition of resemblance remains, and the character of our resistance to their appeals, for we do resist them, is experientially similar. We feel them both as suspiciously sentimental, but find it hard to explain; the resemblance can be detected out of the corner of the eye, as it were, but slips away and proves elusive if the mind attempts to focus upon the matter.

The definition offered above puts content to one side and suggests that the similarity of our sceptical reaction, our attribution of sentimentality, to both the Bodice Ripper and the Bible is explained by the fact that both make an appeal to overwhelming significance without sufficient evidence.

Furthermore, while there can be no doubt that religions aim at transcendence, it is obvious that not all romantic literature goes as far and indeed most seems to stop some way short of this ultimate appeal. Yet who would deny that religion and romantic literature are felt as experientially comparable and in some way similar in structure? Indeed, the matter is notorious.

Again, this powerful yet indefinite intuition of resemblance is explained by an abstract similarity, namely that both make claims to an extreme significance, sometimes but not always in the same terms, that is not supported by robust evidence. In other words, sentimentality is a palpable exaggeration of a degree of significance. More explicitly: the strength of our intuition of sentimentality is explained by the degree of exaggeration. In other words, our intuition that propositions or entire texts are more or less sentimental is determined by variations in the gulf between the evidence and the imputed significance, not the level of imputed significance itself.

This latter point is of the highest importance since it describes the otherwise intractable but well known observation that many of the most deeply sentimental propositions or artefacts make claims to significance that are only modest in absolute terms. For example, expressions of sympathy, observations on the altruism or naive charm of pets, the beauty of flowers, or, and this is a phenomenon novel in our times, the aesthetic experience of food. There may be a hint of an appeal to transcendence in such claims, but frequently there is none. Notwithstanding this we know them to be psychologically cognate to such extreme claims; they are definitely sentimental, and may even be the most nauseating examples of sentimentality, and they are so because the evidence supporting the claim is so weak not because the claim itself is of so great a magnitude. The extravagance is measured by the disparity between the justification and the claim, not the absolute magnitude of the claim.