The role of the dispute about the nature of content-figurae in structural semantics in semiotics’ history: Hjelmslev, Jakobson, and Eco

ABSTRACT

The debate on the semantic analysis of both lexical and grammatical forms has been salient since the early days of the institutionalisation of semiotics, as fundamental to the theoretical and methodological curriculum of this disciplinary domain. In this debate, the concept of content-figurae, or semantic traits, plays a major role. The well-established sources for this concept and the corresponding method of analysis are the works of Hjelmslev and Jakobson, mediated through Umberto Eco’s thinking. In Eco’s framework, the idea of content-figurae is interpreted as that of ‘semantic primitives, and on that ground, its epistemological foundation and methodological efficacy are refuted. In this article, I try to relate Eco’s semiotic position to Hjelmslev’s and Jakobsons’s positions in linguistic (grammatical) semantics, discussing their principles of semantic analysis of grammatical forms from the perspective of the institutionalisation of semiotics as a disciplinary domain.

KEYWORDS:

• Eco
• Hjelmslev
• Jakobson
• semantics
• semiotics

1. Eco and the institutionalisation of semiotics

From its very beginnings as a well-established discipline, around the first half of the 1960s, semiotics has taken Hjelmslev and Jakobson as two pivotal figures for the establishment of its conceptual tools. This process also entailed a certain degree of distancing from the French semiological tradition, which was carried out by Umberto Eco, who gave up “[…] the ‘French’ project of a science des signes closely connected to linguistics and the term ‘semiology’ that belongs to it” (Trabant 2017, 106). Since La Struttura Assente (Eco 1968), he put forward a science of signs based also on the philosophy of Charles S. Peirce”, combining it with some core tenets of Hjelmslev’s and Jakobson’s structural theories of signs. One of these principles is the arbitrary nature of the linguistic sign, which Jakobson borrows from Benveniste’s somewhat problematic interpretation of the Saussurean thesis (Picciuolo 2017). He refers to Benveniste himself in the lecture “Sign and system of language” given on 2 October 1959, in Erfurt:

Of the two fundamental principles of the Cours [author's note: Saussure 1972], les deux principes généraux, as Saussure labelled them, one may see today the first basic proposition – l’arbitraire du signe, the “arbitrariness” of the language sign – as an arbitrary principle. As Benveniste [author's note: 1939] has shown it beautifully in Acta Linguistica 1, from the synchronic point of view of a language community using the language signs, one must not ascribe to them an arbitrary nature. It is not at all arbitrary but rather obligatory to say fromage for “cheese” in French, and to say “cheese” in English. I believe that one may conclude from the whole discussion on “arbitrariness” and “unmotivated signs”, that l’arbitraire was a most unfortunate term choice. (Jakobson and Hrushovski 1980, 33)

This text was chosen as one of the theoretical flagships for the dawning discipline of semiotics: it was published in Italian in an anthology, Lo sviluppo della semiotica, edited and introduced by Umberto Eco in 1978, and then republished in 2020 with an afterword by Nunzio La Fauci, who highlights Eco’s role in mediating Jakobson’s thought:

Umberto Eco’s opening, Jakobson’s semiotic thought, officially serves as an introduction to the volume. Actually, it also represents its main course: it is, in fact, the longest contribution of the book, covering a third of it. With minor adjustments, it is the article that Eco published in 1976 in the miscellany of studies offered to Jakobson on the occasion of his eightieth birthday, “The influence of Roman Jakobson on the development of semiotics”, translated into Italian by La Porta with a new title, “Il pensiero semiotico di Jakobson”. Briefly speaking, in 1974, Jakobson opened the congress of semioticians with a paper giving a short overview of the development of semiotics. In 1976, as a tribute to Jakobson, Eco shows how Jakobson contributed to such a development. In 1978, those essays, along with a few others, became the central contributions to a book bearing Jakobson’s name as its author and mentioning semiotics in the title. This fills a gap in Jakobson’s bibliography, which Eco could not help underlining in 1976, describing Jakobson’s contribution to semiotics. (La Fauci 2020, 144; my translation)

Now, while Roman Jakobson put a strong emphasis on the motivation of any linguistic sign, the other forefather of semiotics considered here, Louis Hjelmslev, highlighted its arbitrary constitution: signs are made up of smaller parts, figurae, which are said to be devoid of any natural denomination, and purely formal (Hjelmslev 1961, 46–47, 67 ff.). Given the centrality reserved by Eco to Hjelmslev’s componential analysis, one might wonder about the consistency of combining the two perspectives at hand into a single semiotic framework. It is significant that Algirdas J. Greimas, another important figure in the institutionalisation process of semiotics, rejects Hjelmslev’s approach precisely on this point, favouring Jakobson’s instead:

Tout mon savoir vient des Prolégomènes. C’est là qui vient la première tentative de Hjelmslev de faire de la sémantique disant que le lexème vache, est égal à « bœuf + féminin ». Sur ce plan je suis Jakobson et non pas Hjelmslev. J’ai suivi ce dernier dans les grands projets, comme la construction de l’univers sémantique. (Greimas 1986, 47)

What reasons led Eco to stubbornly insist – not without some critical distancing – on Hjelmslev’s analysis? In order to understand the role played by Jakobson and Hjelmslev in the construction of the interdisciplinary corpus of semiotics, we will select two points of view. The first is the role played by Umberto Eco in reconstructing the tenets of both Jakobson’s and Hjelmslev’s framework. The second is Eco’s interpretation of the methodological nature and analytical function of content figurae, namely a detail in his argumentation. The detail in question is only apparently so, since – as we will see – it led Eco to carry out an extensive philological analysis of the very texts dealing with the componential analysis of signs into (content) figurae.

2. Semiotics and semantics on content-figurae

Hjelmslev’s semantic claims are discussed in one of the first decisive contributions to semiotics, Roland Barthes’ Élements de sémiologie (Barthes 1964). In this work, reference is made to his formal analysis of linguistic meaning, namely to the way to extrapolate the form of the content of linguistic elements, showing to what extent Barthes was influenced by the Danish linguist on this point:

Hjelmslev, par exemple, décompose un monème comme « jument » en deux unités de sens plus petites: « cheval » + « femelle », unités qui peuvent commuter et servir à reconstituer par conséquent des monèmes nouveaux (« porc » + « femelle » = « truie », « cheval » + « mâle » = « étalon »). (Barthes 1964, 108)

The methodology behind the identification of pertinent semantic features is, as Barthes correctly recalls, commutation: relevant features are those traits of what Barthes calls a given “moneme” that are reciprocally commutable.

Given the importance of Barthes, especially in the dissemination of Hjelmslev’s thought, it is no wonder that, in the introduction to Lo sviluppo della semiotica (Jakobson 1978), Eco points to the Éléments de sémiologie (Barthes 1964) as the first step in the process leading to an international acknowledgement of semiotics as discipline. Yet, he also recognised that this very first step was in turn made possible thanks to the “catalysing” work of Roman Jakobson. The contribution of Jakobson made it possible to distill, so to speak, more than two thousand years of Western thought on the notion of “sign”, giving to it a theoretical consistency that it lacked, fully realising the “transplant” from linguistics to semiotics of the conceptual tools that were needed by this new discipline. Among the many important disciplinary transplants made possible by Jakobson, Eco mentions:

The extension, more or less generalised, of the concept of distinctive feature, closely linked to that of binarism, to numerous other phenomena of signification. Jakobson himself proposed to apply this principle to grammar (1932b, 1936a, 1959a), and I personally believe that this proposal deserves further attention, since it has allowed many semanticists to reformulate their analyses. (Eco 1977, 46)

The works by Jakobson that Eco quotes in this passage are among his most influential in grammatical semantics, namely Zur Struktur des russischen Verbums (Jakobson 1932), Beitrag zur allgemeinen Kasuslehre (Jakobson 1936, 1971) and Boas’ view of grammatical meaning (Jakobson 1959, 1971). It is quite significant that three major contributions in grammatical semantics, thus belonging to the domain of linguistics, are elected as methodological sources for semiotics. These works are not purely theoretical stances on meaning: they are applied studies on a very specific object within grammar: the category of case. Precisely at this point, the views of Jakobson and of his contender Hjelmslev came to clash, for according to Jakobson “meaning is organised in terms of binary oppositions – here called ‘correlations’ – defined by abstract semantic concepts (for example, directionality, shaping)” (Waugh and Monville-Burston 1990, 333). In his review of Hjelmslev’s seminal work on case (Hjelmslev [1935-1937] 1972), Jakobson criticises his Danish colleague for having adopted (quite inconsistently, according to him) a dated stance, despite Hjelmslev’s apparent endorsement of structural principles:

Hjelmslev acknowledges these principles: ‘The structure of the linguistic system does not allow the distinction between a positive term and a negative term to be maintained. The real and universal opposition is that between a definite term and an indefinite one’ […]. But in his descriptions of individual case systems, for example that of the Gothic substantives, Hjelmslev deviates from the above guideline […] Here (e.g., the study of nouns in Gothic), the problem of general meaning is clearly pushed aside, on the one hand, in favour of the traditional list of individual meanings, or of the list of syntactic functions of each of the two cases (for example nominative as the case of the subject and of the predicate, as a predicateless form, and as an address form) and, on the other hand, in favour of establishing the principal meaning (Hauptbedeutung) of each case (in the nominative “the value of ‘subject’ predominates and is often the only one envisaged”), although the author condemns such a procedure in principle (p. 6 and elsewhere). (Jakobson 1990, 340)

This passage sheds light on the reception of Hjelmslev in semiotics. The part of semiotics that has taken up Hjelmslev’s notion of content figurae has pointed towards an ontological relevance of those content-elements: since they belong to grammar, they are systematic and thus more “fundamental” than (lexical) signs. In the passage above, however, Jakobson seems to blame Hjelmslev for the opposite: in Jakobson’s interpretation, Hjelmslev’s fundamental meaning of a grammatical form is neither systematic nor fundamental at all, but rather simple fragments of content, i.e., “individual meanings”, consequently lacking structural soundness.

Despite the caveat coming from Jakobson, Hjelmslev’s semantic description, methodologically redefined in 1943, was fully subsumed in semiotics, diffused, and legitimised by several manuals (see for instance Allan 2015; Geeraerts 2013) as a prototypical case of a componential analysis, also called a “dictionary semantics” approach. And again, despite the precautionary calls by Jakobson (1936) and Barthes (1964), the analysis of the meaning of lexemes (words) advanced by Hjelmslev is received as a procedure based on the discovery of distinctive traits, or “semantic primitives”. An example of this is Violi’s interpretation (2000, 104), in which the analysis per content figurae sketched by Hjelmslev (1943) is taken as a clear instantiation of a method for identifying “primitive” traits – an interpretation the author hold to wrong because it is impossible to distinguish between “features that are really necessary and sufficient to define the meaning of a word and those that are not”. Violi’s interpretation actually rests on the authoritative criticism of Hjelmslevian semantics put forward by Eco on several occasions. This will be discussed in the next section.

3. Content figurae: Eco

The principles of Hjelmslev’s semantic analysis were first quoted by Eco in La struttura assente, published in 1968, four years after Barthes’ Éléments and a year before the establishment of the International Association for Semiotic Studies. In this context, the target of Eco’s critical scrutiny is not the concept of content figurae, but rather the claim about the possibility of defining “the semantic value of a term via the difference in the semantic space occupied by another term”. In fact, he says, “it is hard to see how to establish that the space occupied by the French bois is wider than the space occupied by Italian bosco (if not by already knowing that when I say bosco in Italian, I exclude firewood and timber)” (Eco 1968, 365; my translation). The reference to the same framework, however, also returns in two later major contributions, namely The Forms of Content (Eco 1971) and A Theory of Semiotics (Eco 1976), the latter being a work in which Eco claims to have reworked Hjelmslev’s theses, although rather freely:

Hjelmslev (1943) proposed the possibility of explaining and describing an unlimited number of content entities by making use of a limited number of content figurae, i.e., more universal combinatory features. Given four elementary features such as “ovine” and “porcine”, and “male” and “female”, it is possible to combine them into the sememes “ram”, “ewe”, “pig” and “sow”, these primary universal features remaining at one’s disposal for further combinations. (Eco 1976, 93)

Eco voices his doubts about this kind of componential analysis in his definition of the sign, under the entry Segno (“sign”) in the Einaudi Encyclopaedia:

It remains to be decided […] whether these content figurae belong to a finite system of metasemantic universals or whether they are linguistic entities that in turn clarify the composition of other linguistic entities. (Eco 1980b, 636; my translation)

This question is answered by Eco in the very same work, under the entry Significato (“meaning”, Eco 1980a). Here, Eco points out what he considers to be the fallacy in Hjelmslev’s approach: the meaning of the word gatto ‘tomcat’ corresponds to the group of content-figurae “animal, feline, domestic, male”, which is supposed to constitute its semantic definition, “formulated in the same language and on the same plane to which the entity itself belongs”. This last claim turns out to be problematic to Eco: if equivalence is formulated “in the same language”, there is then a contradiction with the principles of structuralism to which Hjelmslev also adheres: (i) “a totality consists only of relations”, (ii) “in every system of oppositions, something is defined not by what it is, but by what it is not”. There seems to be thus an irreducible ontological difference between the sign and the figura defined as part of a sign, which undermines the very idea of a semantic definition as equivalence:

If instead of /albero/ <‘tree’> you put ‘tall vegetable’, what do you put instead of ‘vegetable’? Thus Hjelmslev leaves a twofold problem unsolved: i) how to define the figurae so that they are not empty differences and not mere expressions; 2) how to establish a limited inventory of them. (Eco 1980a, 847; my translation)

Eco’s posture towards Hjelmslev’s theory is taken up further a few years later, in Semiotica e filosofia del linguaggio (Eco 1984). He acknowledges the historical relevance of this approach, its merit being to detach the analysis of linguistic content from the sterile domains of both mentalism and psychologism. However, he stressed the inability of such an analysis to account for the operations pertaining to the speaker’s linguistic competence:

If someone starts talking and says/he runs/to me, it is not that I, on the basis of my linguistic competence, merely identify a portion of content represented by the articulation of certain figures such as ‘action + physics + fast + with legs, etc.’ […] As soon as I perceive/he runs/, I predispose myself to a series of expectations, by identifying a content space structured as a block of contextual instructions. (Eco 1984, 35; my translation)

The reference to the competence of the speakers, however, is not the only problematic aspect. Reductionism is yet another, or rather the claim that open inventories of signs can be reduced to closed inventories of figurae (see Eco 1986, 47). To do that, Eco resorts to a comparison between the two passages of Hjelmslev’s componential analysis as given in the English edition of Prolegomena (Hjelmslev 1961) and the Italian one (Hjelmslev 1968) and observes that in the former the members of the closed inventory of the personal pronouns “he” and “she” are registered as figurae, instead of the “male” and “female” elements. Eco acknowledges that the choice of personal pronouns as content figurae would in principle guarantee Hjelmslev’s use of a restricted inventory of selecting traits; however, he declares himself dissatisfied with the solution, because “on morphological grounds one obtains a very poor inventory” (Eco 1984, 76; my translation). But this is precisely the reason why such an analysis was put forward by Hjelmslev himself, as a way to limit the number of semantic criteria grounding them on grammatical basis. The point is that Eco intends to use Hjelmslev’s approach for goals that lie beyond its original scope.

4. Content figurae: Hjelmslev

Eco’s wider scope has been recognised also within his own legacy. For instance, Traini (2013) argues that, while drawing upon Hjelmslev’s representation of lexical analyses, Eco discards (or does not consider) the entirety of Hjelmslev’s conceptual framework, which is not oriented towards a theory of knowledge or of human cognition, but simply towards a cognitively neutral general semiological model. It is therefore useful to summarise the basic tenets of this semiological framework in which the distinction between sign and figurae becomes relevant.

1. In Eco’s interpretation, the status of the content-figurae, their very existence, is often questioned within a framework that has previously been defined here as ontological. Quite on the contrary, the existence of an element is for Hjelmslev an attribute that follows from its susceptibility to analysis: an element exists, from a scientific point of view, when it is described (e.g., Fischer-Jørgensen 1966, 17).

2. Despite Barthes’ seminal contribution (1964), which preceded the canonisation of semiotics as a discipline, content figurae are treated as distinctive traits. While Hjelmslev (1943, 65) welcomes the distinctive criterion as part of the framework of glossematic analysis, he does so in a specific way, by reformulating it as a procedure of reduction of larger linguistic elements into content- and expression-invariants in adherence with his ideas of the deductive nature of description and of the hierarchical structure of sign function. Thus, the idea of “invariants” does not perfectly correspond to the notion of “distinctive feature”: in fact, the traditional analysis in distinctive traits contradicts the principles of formal analysis, since it is based on the identification of substantial properties and not on the description of the network of paradigmatic and syntagmatic relations which are said to identify the positional value of any linguistic elements.

3. Indeed, in Hjelmslev the formal procedure does not aim to the identification of semantic primitives or of universals, but to the description of correlations and relation. The difference between signs and figurae is thus not to be understood in the ontological or gnoseological sense: it is purely operational and linguistics-specific.

4. According to Hjelmslev, the difference between signs and figurae is not a difference of size, as size is not a decisive factor in the definition of an element: two elements of the same size can be registered at two different steps of the analysis procedure, and vice versa.

5. The distinction between signs and figurae in the analysis of meaning is instrumental in rejecting the traditional perspective that assigns a meaning to words and to their combinations, thus focusing on the extrinsic functions of signs (usage). Instead, the formal analysis aims to identify and describe the intrinsic functions of the elements of a text in terms of network of dependencies. Now, the traditional view is the one in which, volens nolens, Eco’s interpretation has to be situated.

6. In this framework, signs are not the only elements provided with meaning: figurae, i.e., sign-parts like roots, derivational and inflectional elements also have a specific, infralinguistic meaning. In this respect, a figura is a purely formal and relational element, a content-particle that draws its value from the (closed) paradigm in which it occurs. Hence, it can be identified operatively and labelled arbitrarily by using a sign or a symbol whatsoever.

7. The felicity condition of structural analysis lies in reducing open classes into closed classes. Bringing formal analysis into the domain of the semantics of a language’s lexicon is, from this point of view, the stress test par excellence, since:

S’il y a domaine où le scepticisme à l’égard du point de vue structural retrouve son véritable champ d’aventure et son vrai terrain de jeu c’est celui du vocabulaire. Par opposition aux phonèmes (au sens large, et aux graphèmes etc.) aussi bien qu’aux morphèmes, les éléments du vocabulaire, les vocables ou mots, ont ceci de particulier d’être nombreux voire même d’un nombre en principe illimité et incalculable. Il y a plus : le vocabulaire est instable, il change constamment, il y a dans un état de langue un va et vient incessant de mots nouveaux qui sont forgés à volonté et selon les besoins et de mots anciens qui tombent en désuétude et disparaissent. Bref, le vocabulaire se présente au premier abord comme la négation même d’un état, d’une stabilité, d’une synchronie, d’une structure. À première vue, le vocabulaire reste capricieux et juste le contraire d’une structure.¹ (Hjelmslev (1957) 1970, 97)

Since the entirety of the lexicon in itself can hardly be said to be organised as a totality, a structural analysis of vocabulary, i.e., via closed inventories, must then be carried out in sub-systems, assuming the structure of grammar as foothold. An analysis of this kind was carried out by Holt (1967), resorting to the grammatical distinction of figurae like roots, morphemes and derivatives:

A substantive like lion-ne evidently includes an explicit feminine derivative. Yet a similar derivative of gender can also be found in the couple vache : taureau […] however the derivative is here hidden somewhere in the pleria; now, since the last couple behaves in the same way as lion : lion-ne, we explain this using the following terminology: lion : lion-ne bears the explicit gender-derivative, while vache : taureau contains the same derivative implicitly. (Holt 1967, 65; my translation)²

5. Concluding remarks

The semantic analysis of the lexicon is thus carried out on the basis of a reduction that resorts to closed grammatical paradigms: once again, it is not a matter of identifying the relevant semantic oppositions for the content of a word, but rather the parts of signs (figurae) which a word consists of via a purely relational perspective. The elements “masculine” and “feminine” are proper figurae, and not signs, only insofar it is possible to identify a closed inventory of gender forms like the personal pronouns “he”, “she”, “it”. Eco (1984) is well aware of the morphological nature of the content figurae used by Hjelmslev for the semantic analysis (Eco 2000, 201); nevertheless, his “free” rendering of Hjelmslev’s perspective does not seem to draw the due consequences and pushes it away from its proper premises. Transferred (and adapted) into the domain of semiotics, the notion of content figurae loses a large part of its original complexity.

Consider, for instance, a possible replication of Hjelmslev’s analysis applied to a portion of the lexicon in two stages of a given language, like Italian. Let us take a domain of the Italian lexicon that has been highly discussed from a sociolinguistic perspective: the so-called gender-neutral language and, more specifically, the lexicon of professions; and let put forward the following example, freely reformulated from Hjelmslev (1961). Until not so long ago, the so-called “extended masculine” was current in Italian for many professional nouns, which neutralised, as far as grammatical gender was concerned, the distinction between masculine and feminine. The extended masculine functioned, one might say, as the unmarked element of the gender-paradigm. Thus, taking the pair “sindaco/sindaca” (Eng.: mayor/mayoress), one could observe its functional behavior in two stages. At stage A, we can find the following examples³

Table

(1)a    Lui è il sindaco

‘He is the mayor’

Table

(1)b    Lei è il sindaco

‘She is the mayor’

If one takes “sindaco” as a sign, therefore, both the pronouns “lui” and “lei” enter into a paradigmatic relationship with it, so that the two possibilities described are both in place:

Table

(2)a    Chi è lui? Il sindaco

‘Who is he? The mayor’

Table

(2)b    Chi è lei? Il sindaco

‘Who is she? The mayor’

Table

(2)c    Chi è il sindaco? Lui

‘Who is the mayor? Him’

Table

(2)d    Chi è il sindaco? Lei

‘Who is the mayor? Her’

Thus, at the stage A of Italian, where the morpheme does not bear any explicit grammatical gender marking,⁴ and where the syntactic phenomenon of agreement is neutralised, the two content units: “lui-sindaco” and “lei-sindaco” would be registered. At stage B, however, the following occurrences can be found:

Table

(3)a    Lui è il sindaco

‘He is the mayor’

Table

(3)b    Lei è la sindaca

‘She is the mayor’

giving rise to the following possibilities:

Table

(4)a    Chi è lui? Il sindaco

‘Who is he? The mayor’

Table

(4)b    Chi è il sindaco? Lui

‘Who is the mayor? Him’

Table

(4)c    *(Chi è lei? Il sindaco)

‘Who is she? The mayor’

Table

(4)d    *(Chi è il sindaco? Lei)

‘Who is the mayor? She is the mayor’

Table

(4)e    Chi è lei? La sindaca

‘Who is she? She is the mayoress’

Table

(4)f    Chi è la sindaca? Lei

‘Who is the mayoress? She is the mayoress’

At stage B, considering the sign “sindaco”, the gender-morpheme is no longer implicit, but explicit. The neutralisation of grammatical gender is no longer registered: the phenomenon of agreement is again a relevant feature and the procedure yields different content units than at stage A, namely “sindaca”, “sindaco”, “lei” and “lui” (as personal pronouns). The analysis of this case actually resorts to another paradigm, that of interrogative pronouns (see Hjelmslev (1957) 1970): using this restricted grammatical gender paradigm allows certain productive operations on the semantics of the lexicon, on a par with those performed by operating with personal pronouns. Both are in fact resources that languages use along with other morphological material to mark a series of distinctions in the content, i.e., the shared cultural representations conveyed by a given language, called “collective appreciations” (Hjelmslev (1954) 1970, 52 ff.), which include the fundamental meanings encoded in grammar, like for instance the ideas of relation (case), of consistency (gender and number), of reality (mode), and so on (see Hjelmslev (1938) 1970, 157). This seems to be the challenge posed by Hjelmslev’s structural semantics: to study the “collective appreciations” through the study of the form of linguistic signs (Cigana 2016, 29). Within this framework, the basic idea is to grasp meaning cumulatively: not by approaching it from all perspectives simultaneously, but starting from the core, i.e., grammatical meaning, and proceeding outwards, as it were, building on features that have been already “secured”.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ If there is a domain in which the scepticism towards the structural approach finds its true field of action and playing ground, this is certainly the vocabulary. As opposed to phonemes (in the broad sense, and to graphemes etc.) as well as to morphemes, the elements of the vocabulary, the vocables or words, have the particularity of being numerous or even of a number that is in principle unlimited and incalculable. Moreover, the vocabulary is unstable, it changes constantly; in any state of language there is a ceaseless coming and going of new words, which are forged at will and according to needs, and of old words which fall into disuse and disappear. In short, the vocabulary appears at first sight as the very negation of a state, of a stability, of a synchrony, of a structure. At first glance, the vocabulary remains capricious and just the opposite of a structure.

² Holt’s French text: “Un nom tel que lion-ne contient évidemment un dérivatif explicite du féminin. Mais un dérivatif de genre semble être présent également dans la paire vache : taureau […] mais ici le dérivatif se cache quelque part dans la plérie; or, puisque la dernière paire se comporte comme lion : lion-ne nous nous exprimons à l’aide de la terminologie que voici : lion : lion-ne a le dérivatif explicite de genre, tandis que vache : taureau contient ce dérivatif-ci implicitement”.

³ The analysis applies to Italian expressions; the offered translations are for explanatory purposes only.

⁴ The ending in -o of “sindaco” could then be seen as a syncretism (fusion) of masculine and feminine.