In many languages, the patient argument of a typical transitive clause is coded in some overt way (i.e. not just by word order) to distinguish it from the agent. Such markers are well-known from Latin and are typically called accusative markers (cf. 1).
Accusative markers can be case affixes, as in Latin, but they can also be prepositions, as in many of the examples below, in (7a-c) for example.
As will be seen in Chapter 58, some languages have overt coding of the transitive agent instead, but in this chapter we only consider coding of patients. Moreover, we do not consider patient coding by means of person-number indexing (or cross-referencing, or agreement) on the verb, as is found in Río de la Plata Spanish (cf. 2) and many Bantu languages such as Swahili (cf. 3). (Note that Spanish has both an object person marker on the verb and an accusative marker on the noun phrase.)
Finally, we do not consider the coding of personal pronouns here. Many languages treat full noun phrases and personal pronouns differently (typically restricting accusative marking to pronouns, as in English), and pronouns will be dealt with in Chapter 59.
Thus, for this feature we only ask whether full noun phrases (i.e. nonpronominal noun phrases) have an overt accusative case affix or adposition. This marker need not be unique to patients (cf. Spanish a in (2), which also codes recipients in ditransitive constructions), but it must serve to distinguish patients from agents.
Overt patient marking is not common in APiCS languages.
|No marking of patient NPs||62|
|Only definite patient NPs are marked||2|
|Only animate patient NPs are marked||8|
|Only definite and animate patient NPs are marked||2|
|All patient NPs are marked||2|
Most of our languages have the first feature value, no marking of patient NPs.
This is expected for three reasons: (i) The languages also tend to have fixed SVO word order (cf. Chapter 1), so that the patient NP need not be overtly marked (see Sinnemäki 2010); (ii) the European lexifiers English, French, Portuguese and Dutch do not have overt patient marking; and (iii) hardly any of the relevant African and Oceanic substrate languages have overt patient marking. Thus, overt patient marking is generally found only in Asian pidgins and creoles as well as in some mixed languages.
In the world's languages, accusative marking of patients is not uncommon, but most languages with accusative marking use this marking only on a subclass of prominent patients, especially animate and definite patients. This situation, known as differential object marking (Bossong 1998, Aissen 2003), is more common than consistent object marking, and this is reflected in our languages as well. Moreover, in many cases object marking is optional and its use depends on a complex set of factors. Thus, when we say that only patients of a certain type are marked, this does not imply that these patients must be marked, or that marking of other patients is categorically excluded.
In Sri Lankan Malay and Pidgin Hawaiian, only definite patients are marked (value 2). In both languages, there is variability, but examples (5) and (6) show that accusative marking of inanimate patients is at least possible when they are definite.
In Pidgin Hawaiian, object marking with i (inherited from its lexifier Hawaiian) is in fact uncommon, but when the object is indefinite, it does not occur at all. In Sri Lankan Malay, the accusative marker probably arose under the influence of Tamil. Classical Malay has no accusative marking, and the patient marker sama that is sometimes used in colloquial Malay is not likely to have played a role.
In a number of languages in South Asia and Southeast Asia, only animate patients are marked (value 3). A few examples are given in (7a-c).
In the Spanish- and Portuguese-based languages, the accusative marker typically derives from the preposition con/com 'with'. In Diu Indo-Portuguese and in Sri Lanka Portuguese, the accusative marker is -pa/-pə, from Portuguese para or por.
In the mixed language Michif, patient marking is restricted to animates, too, but the marker in question is traditionally called obviative rather than accusative, because strictly speaking, it marks non-topic status rather than patienthood. Under certain circumstances, it may be used on arguments other than patients. It is counted as a patient marker here because in practice, it mostly occurs on patients.
In Afrikaans and Papiá Kristang, only patients that are both definite and animate get accusative marking (value 4).
Finally, in two languages all patient noun phrases are (or may be) marked with accusative case (value 5). In Zamboanga Chabacano, accusative marking is obligatory on patients that are definite and human, but optionally possible for all patients.
In the mixed language Media Lengua, accusative marking is as ubiquitous as in the grammatifier Quechua, even though it is not absolutely necessary in all cases.