A cognitive semantic approach
to structural focus in the Hungarian clause

Gábor Tolcsvai Nagy
I am grateful for the help and extensive comments of Nóra Kugler and Mária Ladányi on an earlier version of the paper. The research was supported by OTKA (Hungarian Scientific Research Fund), grant K 100717. Certainly, all responsibility is mine.
1The linguistic categories of structural focus and topic have been investigated by formal linguistics in detail. This theoretical and methodological framework defines the two categories as structural positions, filled in by logical operations (cf. É. Kiss 2008), or as factors in information structure (see Lambrecht 1994, Krifka 2008). In what follows, I give a functional cognitive semantic draft of the structural focus, concentrating on the contrastive focus (I do not deal with more specific subclasses of focus). The present interpretation is functional in so far as it takes the perspectival nature of the construal processes of the speaker and the hearer as fundamental and cognitive in so far as it includes the semantic construal processes of the windowing of attention and their results, besides the conceptual elaboration.

In every linguistic interaction, the speaker directs the hearer’s attention (with her or his own) to some entity, from a definite perspective, i.e., the speaker elaborates the entity conceptually from a specific point of view. The processing of the linguistic units (clauses, discourses) is completed in every moment in a restricted conceptual domain, specifying the scope in which something can be conceptualized (Langacker 1987: 118–120, 257–262, 2001: 144, Talmy 2000: 257–309). The conceptual domain attained currently in such a way functions as a complex mental structure, like a window in visual perception, through which the conceptualizer perceives one part of the world, focusing on one thing or event in relation to some entities in the background. To go on with the analogy: things change, and events happen within the frame of the window. Moreover, even the window itself, the attentional frame changes. These circumstances determine how individual concepts are constrained in one moment of processing, and help to designate the entity in the focus of attention.

The focusing of attention is completed in the interaction of the speak­er’s and the hearer’s current perspective. Factors of the interaction are the currently activated knowledge or ignorance, the predictable or diverse, opposite knowledge of the speaker and the hearer. Concentrating only on thing-like entities: the thing in the focus of attention in a wider scope (at least within a clause) functions as topic, with previous activation (with referential accessibility through anaphora), and as reference point (with thematic importance and cataphoric nature) or with inherent topicality (according to cognitive hierarchies) (see Givón 2001: II/226, Langacker 1987, as well as Silverstein 1981, Taylor 1996). Structural focus and contrastive focus in particular arises within the focusing of attention from the relation of the speaker to the hearer’s expectations, the predictability or unpredictability of her/his current knowledge, always in a current situation, in a discourse universe. The speaker and the hearer, while changing roles, jointly form the result of the focusing of attention, i.e., the concentration on an entity, in a negotiation process.

2Linguistic interaction aims at communicating content, at opening up portions of human sense, in joint attentional events. These processes are completed via linguistic structures.

The privileged linguistic structure of communicating content is the sentence, or with a more functional term: clause. The presupposed notion of clause comprises a semantic structure, a phonologic structure, and a more abstract, more schematic morpho-syntactic structure. The parallel correspondences of these structures give the essence of the clause in the functional models. The complexity of the clause can be described in diverse ways, e.g., focusing on the syntactic structure with an abstract semantic content (as in construction grammar; cf. Goldberg 2006, Croft 2001), obtaining clause from the semantic structure (as in Langacker’s cognitive grammar; cf. Langacker 1987, 1991, 2009), concentrating on the correspondences between semantics and syntax (and also pragmatics, as in Givón 2001).

A clause never stands “alone”; it functions as a structure in a processed situation (in a current discourse space, a discourse universe). In the default case the basic structure is the clause (and not sentence), a unit of spontaneous discourse as well as planned written texts.

The prototypical clause has the following features (see Langacker 2008: 354–405). On the semantic pole it maps an event (a scene), with two participants designated. Participants are designated prototypically by nouns conceptualizing things. The clause maps the relation between the two participants temporally. Temporal relations are expressed by verbs, designating an interaction with energy transfer or change (in the physical or metaphoric sense) between two or more participants.

The event expressed in the clause is construed by a conceptualizer, particularly the speaker, from a specific perspective that directs the speak­er’s and the hearer’s attention. Certain linguistic expressions focus the speak­er’s and the hearer’s attention on particular things and their relations in the discourse universe. One component of construal is focusing in a wider conceptual domain (or in domains), the selection of the content to be expressed linguistically, in foreground — background relations.

The speaker and the hearer focus on one of the semantic (conceptual) components of the clause, most often one of the named participants. The attention is focused even in a clause with neutral intonation and stress:

A tűz kialudt.
‘The fire died out.’
A portás becsukta a kaput.
‘The porter closed the gate.’
Az igazgatónak tetszik a titkárnő.
‘The director likes the secretary.’
Jánost érdekli a relativitáselmélet.
‘John is interested in the theory of relativity.’

In (1)–(4) attention is directed on one particular participant (underlined in the examples), through specific cognitive processes, above all by prominence and accessibility hierarchies. Prominence and accessibility contribute to perspectivization, to the designation of the one participant that is elaborated conceptually by the others (cf. Langacker 1987: 120–132, 2008: 55–85, Talmy 2000: 257–309, 311–343, Taylor 1996). The entities placed in the focus of attention by the speaker in (1)–(4) are easily accessible, more easily so than the other participants (if there are any), because they are in a front position on the prominence and accessibility hierarchies, compared to the other ones. This way of attention focusing is not altered, only completed by flexible word order in Hungarian, in (1)–(4) with sentence initial position. In other words: in Hungarian declarative clauses without a structural focus, the word order position of the entity in the focus of attention does not (or does not necessarily) influence its attentional status.

In what follows, I apply the functional cognitive framework outlined above in the interpretation of structural focus in the Hungarian language. The important results of the formal investigations of Hungarian are based on the idea of autonomous syntax and truth conditional semantics (see, among others, É. Kiss 1998, 2002, 2008, Kenesei 1998, in a different manner Kiefer & Gyuris 2006, Kálmán 2001). For the first functional cognitive approaches to the Hungarian language see Imrényi (2008).

3Attention is focused prototypically on a participant conceptualized as a thing and expressed by a noun as a default. Clauses in natural discourse usually elaborate one new information unit (cf. Givón 2001: II/222, Chafe 1976, 1994: 108–120). One characteristic of the semantic structure of the linguistic expressions (e.g., words) plays an important role in the windowing of attention. The semantic structure of the concept of things is processed in its immediate scope in a clause, not in its maximal scope (cf. Langacker 1987: 118–120). For instance, the immediate scope of the noun hand is arm, its maximal scope is body. Also, the semantic content of the noun is bounded, constrained: as a part of its semantic structure, its substructures are abstracted concepts forming a network (a complex matrix), thus a specific noun as a category and as an epistemically grounded instantiation, too, is delimited from other concepts. In (1)–(4), the epistemically grounded concepts of things in the focus of attention comprise their separation from other concepts as a default, without any specific marking. This detachment is not specific, i.e., it does not contain a spreading activation that would determine the conceptual domain (or domains) to be activated by some kind of motivation attached to the concept in the focus of attention (see Deane 1992), or if it contains some element of association, the conceptual delimitation does not need distinct marking.

In these cases the level of presupposition, assertion, accessibility and informational probability of the focused participant in the clause is high both for the speaker and the hearer within the current discourse space. In other words, the speaker’s expectations about the hearer’s knowledge (almost) agree with the hearer’s knowledge (or ignorance). This is the neutral, normative context (Givón 2001: II/223).

In other cases the conceptual delimitation does need distinct marking: the conceptual delimitation of the participant in the focus of attention should be designated within the semantic structure of the clause, to separate it from other participants of the events. These participants are the individuals associated contextually, in a conceptual domain conceptualized by the speaker from a specific perspective. In (5a), identical to (2), both the contextual conceptual detachment of portás ‘porter’ and kapu ‘gate’ is sufficient without any marking, because the speaker’s expectations about the hearer’s knowledge agree with the hearer’s knowledge, according to the speaker’s judgement. This does not hold in the case of (5b) and (5c):

  1. A portás becsukta a kaput.
    ‘The porter closed the gate.’
  2. A PORTÁS csukta be a kaput (és nem a sofőr).
    ‘It was the porter who closed the gate (and not the chauffer).’
  3. A KAPUT csukta be a portás (és nem az ablakot).
    ‘It was the gate the porter closed (and not the window).’

In (5b) kapu ‘gate’ is an activated and bounded participant of the event expressed in the clause both for the speaker and the hearer within the current discourse space, but the agent of the action is not: for the speaker sofőr ‘chauffer’ is the agent, for the hearer (also a speaker in a dialogue) it is the portás ‘porter’ who acts. In (5c) portás ‘porter’ is an activated and bounded participant of the event expressed in the clause both for the speaker and the hearer within the current discourse space, but the patient of the action is not: for the speaker ablak ‘window’ is the patient, for the hearer (also a speaker in a dialogue) it is the kapu ‘gate’ that suffers the act. That is, in both cases of (5b) and (5c) there is a contrast between the activated knowledge and expectations of the speakers and hearers. They designate this contrast through the focusing of attention, by conceptualizing the focused entity in relation to another entity, according to their two diverse activated segments of knowledge. Conceptualization takes place within the processing of the discourse universe, the current discourse space, based on the relation between the current hearer’s expectations and the predictability of the current speaker’s knowledge. A simple dialogue gives a clearer picture of the perspectives and attention directing of the speakers and hearers:

A: János egy órakor megérkezett.
   ‘John has arrived at one o’clock’
B: Nem, egykor ZSUZSA érkezett meg.
   ‘No, it was Susan who arrived at one o’clock.’

There is a contrast between the knowledge portions activated by the two speakers in dialogue (6). Both of them speak about a series of events with two participants, wherein one event is connected to an exact date. The participants of the event are known to the interlocutors, as is the event (the arrival). Only the link between the right participant and the right date differs for the two interlocutors. For interlocutor A participant János arrived at the mentioned time (1 pm, in the first turn). The attention focusing on this participant (János) can be completed with the background of concepts already activated in the previous sections of the discourse and discourse space, without particular designation, the conceptual delimitation needs no specification, epistemic grounding is sufficient. In the second turn of the dialogue, interlocutor B construes a contrast: according to her/his knowledge it was Zsuzsa who arrived at 1 pm. The participant named Zsuzsa was already activated in the discourse or the discourse universe. The conceptual scoping of the participant named Zsuzsa in the discourse is possible only in contrast with the other participant (János) mentioned in the first turn of the dialogue.

The conceptual delimitation is explicated here: from a contextually, within the current discourse space construed or schematically activated category [it is X and not Y], in a more general way: [X in contrast with Y and Z]. This clause internal conceptualization is stressed and explicit, because it is not part of the presuppositions, expectations and information predictability derivable from the comprehension processes of one of the speakers (speaker B) in advance, in contrast to the other speaker (speaker A). In the Hungarian language the semantic structure focused in that way is designated prototypically by main clausal stress and preverbal position, and the postverbal position of the elements categorized as verbal modifiers. Nevertheless, this schema should be investigated on corpus data to be validated, and to find out the real spectrum of syntactic and semantic variability. It has to be noted that stressed focusing can be completed also on grammatical elements, as in (7):

A: Nem láttad a zseblámpámat?
   ‘Have you seen my flashlight?’
B: A komódban van.
   ‘It is in the chest of drawers.’
A: Nincs a komódon.
   ‘It is not on the chest of drawers.’
B: A komódBAN van, nem a komódON.
   ‘It is IN the chest of drawers and not ON it.’

The focus (structural focus) interpreted this way can be described within context, within the universe of discourse. Part of the discourse universe is the set of concepts, whether elaborated or schematic, activated by the conceptualizers. An example, discussed repeatedly in the Hungarian literature, may be described in such a way:

János a PADLÓN aludt.
‘János slept ON THE FLOOR.’

The speaker in this clause not only presupposes or takes already activated in the given discourse space that there are other participants in the represented events besides János for the conceptualizers. But the clause also contains that floor as the place of sleeping does not belong to the concept of sleeping, rather the opposite. Padló ‘floor’ has to be dissociated and focused as a concept, because it cannot be expected on the basis of the informational predictability arising from general schematized knowledge (the schema of sleeping), above all not from the hearer’s side. The speaker’s schematic knowledge has been modified by the event to be expressed. The construction with structural focus and main stress takes this contrastive character in the foreground in elaborating the concept of János (others slept in other places). Nevertheless, the variability of the interpretation of (8) depends on the context: this clause has a neutral variety whereby the conceptual scope of János prevails in its default version, the conceptual focusing is not designated in particular (e.g., János was alone; still he slept on the floor).

It has to be noted once more that structural focus interpreted in the framework outlined above does not stand on its own, in spite of the conceptual isolation and foregrounding: the linguistic unit functioning as focus is part of the semantic structure of the clause (by partial correspondences, valence relations), and at the same time it is topic (in the cognitive sense; cf. Givón 2001: II/229) as designating an already activated concept in the universe of discourse. The activated status does not need direct lexical naming; it can be the result of a semantic activation, the processing of a schema, a spreading activation based on the previous discourse parts or the processed discourse universe without an overt linguistic expression (cf. Chafe 1994). The prototypical structural focus is built on the contrast between the speaker’s and the hearer’s presuppositions and expectations. What counts as activated unit in the discourse (with topic status) for the speaker, for the hearer may be non-activated, not known or contrary to her/his knowledge, not fitting to her/his expectations. Still, in the default case, the contrastive focus is founded in the previous discourse parts, and it is an accessible topic through anaphor.

The cognitive and communicative configuration with the conceptual scoping and delimitation outlined above is extended to other, similar construals. Focusing, like those mentioned before, is related to epistemic grounding, but is not identical to it.

4Structural focus has various instantiations in Hungarian, too (cf. Givón 2001: II/224):

a. Main stress + word order

A FIÚ csukta be az ablakot.
‘It was the boy who closed the window.’
Az ABLAKOT csukta be a fiú.
‘The boy closed the window (as opposed to closing something else).’
BECSUKTA a fiú az ablakot.
‘The boy closed the window (as opposed to doing something else to it).’

b. Cleft

A FIÚ az, aki becsukta az ablakot.
‘It is the boy who closed the window.’
Az ABLAK az, amit becsukott a fiú.
‘It is the window that the boy closed.’

c. Pseudo-cleft

Az a valaki, aki becsukta az ablakot, a FIÚ.
‘The person that closed the window is the boy.’
Az a valami, amit becsukott a fiú, az ABLAK.
‘The thing that the boy closed is the window.’

Version (a) is the least constrained semantically and syntactically, versions (b) and (c) do not include the formation of the verb.

5The syntactically flexible Hungarian word order is based on iconicity: it is the semantically structured sequence of activated concepts, wherein a linguistic unit activated earlier may involve the later ones in the conceptual domain created by spreading activation, opened by itself. The clause initial units have a determining function, although later clausal elements may reinterpret the initial conceptual status. Structural focus is instantiated with a more rigid word order, and a more restricted scope within the clause in the perspective of the windowing of attention.

Word order has a specific semantic function in the instantiation of the preverb + verb or other modifier + verb structures. The preverb profiles the spatial component or one of its metaphorical extensions of the process designated by the verb. The root verb connects the designated temporal process directly to the concept of the thing (a figure) specified in the dominion of the verb, epistemically grounded, maintaining the conceptual foregrounding (cf. Tolcsvai Nagy 2005).

6Summary. The speaker directs the hearer’s and her/his own attention to a clausal component in every clause. The focused thing in the clause is constrained as a default for both the speaker and the hearer, corresponds to their expectations, and is separated from other concepts without specific overt marking.

The focusing of attention has diverse functional instantiations, according to the symmetry or asymmetry of the speaker’s and the hearer’s expectations and presuppositions, the contextual predictability and the degree of focusing.

In the syntactic category identified as structural focus the entity focused conceptually by the speaker does not correspond to the hearer’s expectations and presuppositions, thus the focused entity should be separated and delimited in a contextually determined conceptual domain. The instantiations of focusing are directed by semantic and pragmatic factors.


Chafe, Wallace. 1976. Givenness, contrastiveness, definiteness, subjects, topics and points of view. In: Charles N. Li (ed.), Subject and topic. London: Academic Press. 25–55.

Chafe, Wallace. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago & London: The University of Chicago Press.

Croft, William. 2001. Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press.

Deane, Paul D. 1992. Grammar in Mind and Brain: Explorations in Cognitive Syntax. Berlin & New York: Mouton de Gruyter.

É. Kiss, Katalin. 1998. Identificational Focus versus Information Focus. Language 74: 245–273.

É. Kiss, Katalin. 2002. The Syntax of Hungarian. Cambridge: Cambridge University Press.

É. Kiss, Katalin. 2008. Topic and focus: two structural positions associated with logical functions in the left periphery of the Hungarian sentence. Acta Linguistic Hungarica 55/3–4: 287–296.

Givón, Talmy. 2001. Syntax. A functional-typological introduction. Amsterdam & Philadelphia: John Benjamins.

Goldberg, Adele E. 2006. Constructions at work. The nature of generalization in language. Oxford: Oxford University Press.

Imrényi András. 2008. Szerkezeti fókusz — pragmatikai típusjelölés? Általános Nyelvészeti Tanulmányok XXII: 203–239.

Kálmán László (ed.). 2001. Magyar leíró nyelvtan: Mondattan I. Budapest: Tinta Könyvkiadó.

Kenesei István. 1998. Adjuncts and arguments in VP-focus in Hungarian. Acta Linguistica Hungarica 45: 61–88.

Kiefer Ferenc and Gyuris Beáta. 2006. Szemantika. In: Kiefer Ferenc (ed.), Magyar nyelv. Budapest: Akadémiai Kiadó. 175–221.

Krifka, Manfred. 2008. Basic notions if information structure. Acta Linguistic Hungarica 55/3–4: 243–276.

Lambrecht, Knud. 1994. Information structure and sentence form. Cambridge: Cambridge University Press.

Langacker, Ronald W. 1987. Foundations of Cognitive Grammar. Volume I. Stanford: Stanford University Press.

Langacker, Ronald W. 1991. Foundations of Cognitive Grammar. Volume II. Descriptive Application. Stanford, CA: Stanford University Press.

Langacker, Ronald W. 2001. Discourse in Cognitive Grammar. Cognitive Linguistics 12/2: 143–188.

Langacker, Ronald W. 2008. Cognitive Grammar: A Basic Introduction. Oxford: Oxford University Press.

Langacker, Ronald W. 2009. Investigations in Cognitive Grammar. Berlin & New York: Mouton de Gruyter.

Silverstein, Michael. 1981. Hierarchy of features and ergativity. In: Robert M. W. Dixon (ed.), Grammatical categories in Australian languages. Canberra: Australian Institute of Aboriginal Studies. 227–244.

Talmy, Leonard. 2000. Toward a cognitive semantics. Vol. I: Concept structuring systems. Cambridge, MA: The MIT Press.

Taylor, John R. 1996. Possessives in English: An exploration in cognitive grammar. Oxford: Oxford University Press/Clarendon.

Tolcsvai Nagy Gábor. 2005. Kognitív jelentéstani vázlat az igekötős igéről. Magyar Nyelv CI: 27–43.