Object Alignment in Linguistics | Mini-Lesson

This section examines dative shift as a second major argument alternation alongside passives. Dative shift allows verbs to take either a prepositional phrase (prepositional dative) or a secondary direct object (double object), as in 'gave a bone to the dog' vs. 'gave the dog a bone.' Not all verbs allow dative shift—verbs like 'pull' can take prepositional datives but not double objects. The passive and dative shift interact complexly: the passive of prepositional dative yields a different result than the passive of double object. Implementing these rules requires careful ordering of lexical rules to capture all valid alternations while blocking invalid ones.

English has two dative constructions for giving: the prepositional dative ('Give a muffin to a mouse') and the double-object dative ('Give a mouse a muffin'). Hundreds of verbs can appear in both constructions, leading to a tempting generalization that any verb usable in one form should work in the other. However, there are idiosyncratic exceptions, such as 'Biff drove Chicago the car' being ungrammatical while 'Biff drove the car to Chicago' is fine, and 'Sal gave a headache to Jason' sounding odd compared to 'Sal gave Jason a headache'.

English allows two structures for the dative case: the double object construction (verb + indirect object + direct object) and the prepositional construction (verb + direct object + to + indirect object). For example, 'Give him a book' and 'Give a book to him' both mean the same thing. The double object construction is more common and natural in English. Common verbs that follow this pattern include: send, show, lend, read, write, sell, promise, offer, teach, and cook.

Dative constructions are syntactic structures with an agent, a verb of transfer of possession, and a theme moving to a recipient. In standard Portuguese, these use 'para' (e.g., 'João deu o livro para a mãe'). In Portuguese of Angola, variations include 'na' (e.g., 'João deu o livro na mãe') or 'a' (e.g., 'João deu o livro a mãe'). Cross-linguistically, Latin uses dative case, Irish uses dative constructions, and English presents double object and prepositional dative constructions. This typological perspective helps understand how different languages encode similar semantic relationships.

Irish and Scottish Gaelic are non-alternating languages lacking the double object construction (DOC), using prepositional dative constructions instead. Four theoretical camps exist for analyzing dative alternation: distinct argument structures, locative small clauses, applicability analysis, and derivational relationships. Harley's P-log analysis proposes P-log for prepositional dative (locative semantics) and P-have for DOC (possessive semantics), correctly predicting that languages without P-have cannot have DOC. However, this analysis fails to explain backward binding from indirect object into direct object, which requires the indirect object to c-command the direct object—a position it doesn't occupy in the P-log structure. An alternative analysis derives prepositional dative constructions through VP movement: the direct object raises from P-have complement to become the preposition complement, the indirect object raises to Spec,APPLP, and the VP containing the direct object moves over the indirect object. This resolves locality issues by moving a VP rather than a DP. Irish and Scottish Gaelic are inverse VO languages, showing rightward shift of specific arguments (definites, quantified expressions) rather than leftward shift, which is explained by VP movement: the definite moves to a specifier position, then the evacuated VP moves over it.

Some Italian regional varieties (especially central and southern dialects) use differential object marking with the preposition 'a' for certain objects. For example, 'Ho visto a Maria' (I saw Maria) uses 'a' for human, definite objects, while 'Ho visto un film' (I saw a film) does not. This pattern is also found in Spanish and Sardinian. The marking correlates with semantic traits like animacy and definiteness, and is a legitimate grammatical strategy, not an error.

Differential object marking in Spanish shows animacy relevance—animate objects require case marking while inanimate objects do not. Recent studies propose DOM results from DP movement driven by case requirements. Following Richards (2008), plus animate objects have person features (+person) and must move to check case in a functional projection above VP. Minus animate objects and plurals lack person features and can value case in C. This framework explains animacy restrictions on null objects.

Differential object marking is a grammatical feature where the case assigned to a direct object depends on semantic properties of the object, most commonly animacy. In many languages, inanimate direct objects remain unmarked (in their unmarked case, usually nominative), while animate direct objects take an accusative marker. This is seen in Turkish, Amharic, and Hebrew. The creator applied this to their conlang, where inanimate nouns never take the accusative case marker because they are recoverable from context, while animate nouns require it. This system simplifies grammar by reducing case markers needed for inanimate nouns.

Purépecha exhibits Differential Object Marking (DOM) with nominative-accusative alignment. The suffix 'ni' marks objects of monotransitive and ditransitive verbs. Two DOM patterns exist: (1) Extinct type where intrinsic properties trigger marking (human referents, demonstratives, plural inflection); (2) Fluid type where the same noun phrase may or may not receive marking based on context. The language lacks articles, making case marking crucial for definiteness encoding. Case marking with 'ni' functions as a resource for encoding definiteness, delimiting the set of possible referents. This resource can be omitted when interpretation can be established from linguistic and extralinguistic information. Bare nouns with non-human referents and absence of case marking typically present indefinite readings. Indefinite noun phrases with 'ma' exhibit fluid case alternation: absence of 'ni' yields non-specific readings, while presence yields specific, partitive, or partitive specific readings. Case marking is not required for specific or partitive interpretations when relative clauses or contextual information delimit reference. The partitive specific reading can be inferred from situational context, shared knowledge, and textual information. Case marking with 'ni' occurs with bare nouns having inanimate referents and definite readings, but can be omitted when the referent can be unequivocally recognized. The definite reading can be established through relative clauses, contextual inference, or when the referent is already active in the discourse. This demonstrates that definiteness encoding involves an interaction between grammatical marking and contextual factors.

Differential Object Marking (DOM) in Western Romance languages exhibits microvariation across Spanish, Catalan, Italian, and Portuguese, driven by animacy, referentiality, and effectiveness factors; Spanish and Catalan show the broadest DOM distribution with animacy as the primary trigger, while Portuguese demonstrates the weakest system where only divine/superior nouns and first/second person pronouns require marking, revealing a hierarchy of DOM factors where different nominal and verbal categories are marked according to syntax-semantic principles.

The World Atlas of Language Structures (wals.info) is a typological linguistic database containing information on approximately 500 languages. Unlike Ethnologue's anthropological approach, WALS focuses on cross-linguistic typological features organized into categories such as phonology, morphology, syntax, and word order. Each feature has multiple value categories (e.g., consonant inventory sizes: small, moderately small, average, moderately large, large), with associated maps showing geographic distributions. This tool enables visualization of how linguistic features cluster geographically across different language families.

The World Atlas of Language Structures (WALS) is a comprehensive database mapping grammatical diversity across the world's languages. Originally published as a 700-page book with interactive CD-ROM tools, it has been made freely available online at wals.info. The atlas represents structural diversity through maps showing distributions of features like tone systems, gender systems, word order, and numeral classifiers. This resource enables researchers to visualize and analyze cross-linguistic patterns systematically, revealing geographic distributions of linguistic features and testing hypotheses about language relationships and diffusion.

The World Atlas of Language Structures (WALS) is an online resource created by linguists that provides an overview of language typology. It contains data extracted from grammatical descriptions of individual languages, allowing researchers to compare features across hundreds of languages. According to WALS, SVO (Subject-Verb-Object) is the most common word order pattern globally, followed by SOV (Subject-Object-Verb). VSO (Verb-Subject-Object) is less common, and VOS, OSV, and OVS are very rare. The instructor notes that some linguists argue that certain patterns like OVS are impossible due to cognitive constraints.

In ditransitive constructions (verbs with recipient and theme arguments), the choice between neutral alignment (both objects treated equally) and indirective/cative alignment (one object marked specially) depends on referential properties along four scales: person (1st/2nd > 3rd), definiteness (definite > indefinite), animacy (animate > inanimate), and nominality (pronouns > full noun phrases). A non-aoristic, usage-based approach explains that grammatically frequent constructions show minimal coding (neutral alignment), while rare constructions require additional marking (indirective/cative alignment). This pattern holds universally across languages, as demonstrated by examples from English, French, Czech, Arabic, Wallof, and other languages, where downstream scenarios (recipient more prominent) use neutral alignment while upstream scenarios (theme more prominent) require special coding.