If a material thing is bound to be light, medium heavy (or 'medium light') or heavy, it is bound to have one of the properties of a certain set of properties which are inseparably connected with each other. And if it is bound to be lighter than, as heavy as or heavier than another material thing, it is bound to have one of the relations of a certain set of relations which are inseparably connected with each other. The set of properties is not a primary, simplex thing, for an object cannot be light (have the property of lightness) and medium heavy or heavy (have the property of heaviness) at the same time: it is the extensionality of a secondary thing. The properties in the first set are therefore to be conceived of as secondary things themselves, that is, as simplex things in the second domain of discourse. The relations in the second set can only be things in the second domain of discourse, but also the set of these relations is a mere extensionality belonging to a complex, secondary thing. Neither the set of connected properties nor the set of connected relations is purely conceptual, because then we would be faced again with the 'logical' predicament in which heavier than could turn out to be formally combined not only with as heavy as but also with as slow as or some other equivalence relation. The 'intensionally' connected series of lighter than, as heavy as and heavier than must distinguish itself from a mere set (such as that of heavier than and as slow as), because it is an existing whole with its own secondary predicates. The relationships between its constituent parts are now not merely conceptual or accidental any more: the primary predicates which constitute one and the same whole in the second domain of discourse are linked together or 'concatenate'. The connected series of related primary predicates itself we shall call "a catena", "a catena of attributes or relations".

A predicate catena comprises all the, and only those, primary predicates that are inseparably linked together. Hence, the relations heavier than and as slow as do not constitute a catena, because talk of both relations presupposes the existence of other catenas before we can even speak of "heavier than" and "as slow as". But the relations heavier than and as heavy as do not constitute a catena either, since the relation lighter than is needed to complete its extensionality. (As light as may mean the same as as heavy as, lighter than certainly does not mean the same as heavier than.) The attributes light and heavy do not make up a catena, because they are inseparably connected with an attribute lying between lightness and heaviness: neither light nor heavy (but having a weight nevertheless) (if thinking of a borderline) or medium heavy (if thinking of a transitional zone). The terms light and heavy are so-called 'vague' predicate expressions in that it is not always evident when light becomes heavy, or conversely. It is therefore better to first illustrate the new concept of the catena by means of some other, clearer examples.

The predicates of electropositivity, electroneutrality and electronegativity cannot exist without each other and make up the extensionality of --what we shall name after its positivity-- the 'electropositivity catena' (of which secondary predicates form the attributive predicament). The attributes of happiness, of the state of indifference between happiness and unhappiness and of unhappiness form extensionally the happiness catena. Increase, decrease and the concatenate neither increasing nor decreasing are together the component parts of an increase catena. (There is not one increase catena but there are many of them.) Rest, or being-at-rest, is itself a borderline predicate between motion in positive and motion in negative direction, and therefore rest and motion together constitute a motion catena without any third predicate or set of predicates. Just as increase catena is the common denominator of all increase catenas, so motion is the common denominator of all positive or negative velocities or degrees of motion, happiness the common denominator of all degrees of happiness; and so are unhappiness, electropositivity, and so on.

What is special about heavier than and as heavy as is that they belong to the same catena of relations (together with lighter than). There is a real relationship of having between this one catena and the relation heavier than and also between this catena and the relation as heavy as. The so-called 'relationship' of catenation, that is, of belonging-to-the-same-catena, between heavier than and as heavy as themselves is therefore not ontic. Nevertheless it refers to the existence of a real, but abstract, whole with real, but abstract, components. What deserves our attention in particular is that this conceptual relation of catenation between primary relations is not at all different from the one between attributes as heaviness and lightness or, for that matter, happiness and unhappiness. It would therefore be a rather fragmentary approach to recognize a formal property like connectedness (our 'concatenatedness') only with regard to relations and not with regard to attributes, especially when the existence or recognition of the 'qualitative' attribute catena precedes or is presupposed by the existence or recognition of the comparative relation catena.