The notion of sign appeared rather recently. However, the notions equivalent to the signifier and the signified were discussed in science from the times of the ancient Greeks. For several centuries, it has been debated whether the signifiers of things are given to us by nature or by human convention.
The proponents of the first point of view argued that some words in any language directly arose from onomatopoeia, or sound imitation. Indeed, we denote the sounds produced by a cat with verb mew or meow in English, maullar in Spanish, and miaukat’ in Russian. Hence, according to them, common signifiers lead to similar signifieds, thus creating a material link between the two aspects of the sign.
The opponents of this view objected that only a minority of words in any language takes their origin from such imitation, all the other words being quite arbitrary. For example, there is no resemblance between Spanish pared, English wall, and Russian stena, though all of them signify the same entity. Meanwhile, the phonetic similarity of the German Wand to the English wall, or the French paroi to the Spanish pared, or the Bulgarian stena to the Russian stena are caused by purely historic reasons, i.e., by the common origin of those pairs of languages.
The latter point of view has become prevalent, and now nobody looks for material links between the two sides of linguistic signs. Thus, a linguistic sign is considered a human convention that assigns specific meanings to some material things such as strings of letters, sequences of sounds, pictures in hieroglyphic writing or gestures in sign language.