Türkçenin kimi diller gibi çift nesneli bir dil olduğu ileri sürülür. Dil bilimsel açıdan çift nesneli yapılar üç istemli fiillerin söz dizimsel çevresinde gerçekleşebileceğinden ve bir fiil mantıksal üyelerine aynı anda ancak bir anlam bilimsel rol ve dil bilgisel durum yükleyebileceğinden, söz diziminde aynı rolü yüklenen birden fazla yükleme durum ekli üyenin gerçekleşme olasılığı ilkece mümkün değildir. Bu dil bilimsel görüş, bu çalışmada bir yapının çift nesneli olup olmadığını test edebilmek için kullanılabilecek bazı ölçütlere dönüştürülmüştür: istem bilgisi, rol ölçütü, durum ölçütü ve özneye taşınabilirlik. Diller istem bilgisi, rol ölçütü ve durum ölçütü ilkeleriyle genellikle uyuştuğundan çift nesneli yapıların çift nesneli dillerde bile belirli yapısal dönüşümlerle başka yapılardan türetildiği ileri sürülür. Türkçede çift nesneli olduğu iddia edilen yapılar çift nesneli yapılara benzemekle birlikte hem söz dizimsel hem de anlam bilimsel açıdan gerçek çift nesneli yapılardan farklı özellikler taşır. Bu yapılar İngilizcedeki ne gerçek çift nesneli yapılara ne de ECM yapılarına benzer. Çift nesneli görünen Türkçe yapılar, iç cümlenin öznesinin ana cümlenin nesnesine yükseltilmesiyle oluşan küçük cümleciklere dayanır. Küçük cümleciğin yüklemi ana cümlenin fiiliyle söz dizimsel isim-eylem birleşmesine maruz kalarak yeni bir söz dizimsel oluşum geçirir. Bu söz dizimsel oluşumda küçük cümleciğin yüklemi fiil tamamlayıcısı adı verilen söz dizimsel bir unsura dönüşür. Fiil tamamlayıcılı yapıların ettirgenleştirme işlevi olduğu ve çoğunlukla bilişsel fiillerin söz dizimsel çevresinde gerçekleştiği görülür

It is argued in Turkish syntactic studies that a sentence can have two objects; thus, double-object constructions exist in Turkish too. The structures in (1-7) which are considered to have double-objects actually have different syntactic and semantic features. Although these structures seem to have double objects, they do not meet the syntactic and semantic conditions that a standard double-object construction carries. 1) Her çağ, kendi rüyalarını, kendi emellerini söyletmiş kelimeye. 2) Onu başkan seçtik. 3) O, toprağı altın yaptı. 4) Seni arkadaş sanırdım. 5) Onu vali tayin ettiler. 6) Ahmet Hasan’ı pencereyi açtı sanıyor. 7) Ayşe bu kitabı Ali’ye verdi. Linguistic theories such as Universal Grammar, Case Grammar and Valency Grammar posit that double-object constructions can occur within the syntactic structures of verb arguments with three valencies, and that a verb can assign only one semantic role and grammatical case to its logical arguments at the same time. Thus, as a syntactic rule, it would not be possible for more than one argument with accusative case marking to be assigned with the same role. In this study, this linguistic approach is used as the basis for the following criteria to test whether a structure has double objects or not: valency, role criterion and case criterion. Beside these three criteria, another criterion to test whether a structure is double-object or not is the ability of object to be moved to subject through passivisation. Since languages comply with the three criteria - valency, role and case, it is suggested that double-object constructions even in double-object languages are derived from other structures through certain structural transformations. Though the supposedly double-object constructions in Turkish have similarities with real double-object constructions, the former differs from the latter. The Turkish structures in question are not similar to the real doubleobject constructs in English or ECM structures. In Turkish syntax, only determinate noun phrases and indeterminate noun phrases are possible. These constructs contradict the more strict case criterion in Turkish as well as valency, role criterion and object criterion. Double-object constructions occur within the syntactic and semantic boundaries of a verb with three valencies. The first argument of a verb with three valencies realises as subject in bare case (nominative), the second and third as object-1 and object-2 in accusative case. Real double-object constructs have a semantic role structure that occurs where the arguments of verb are assigned agent+patient+theme. It could be noted that the role structure of the double-object constructs have a meaning that can be defined as “transfer”. This semantic structure indicates that someone (agent) transfers something (theme) to somewhere or someone else (patient). When the supposedly double-object constructions in Turkish are analysed in this line, it is seen that the verbs söylemek and vermek in (1) and (7) have three valencies, while the verbs seçmek, yapmak, sanmak, tayin etmek have two valencies. Although the verbs söylemek and vermek have three valencies, the verb söylemek does not have the syntactic or semantic features of double-object constructions. The verb vermek, however, only complies with double-object constructions in terms of semantic roles. Yet, that the verb vermek consists of the role structure of double-object constructions have is not sufficient to make it a double-object constructions. Although the verbs seçmek, yapmak, sanmak, tayin etmek have two valencies, in (2-6) they behave like they have three valencies by assigning case to determinate and indeterminate objects. However, since a two-valence verb cannot open a valence slot out of its logical valency and that it cannot assign the same case and role to its arguments, the constructs in (2-6) should have been exposed to another syntactic process. It is seen in the literature that structures like (6) in particular are explained through raising and exceptional case assignment. In terms of raising, the Turkish structures in question are considered as the structures where the subject of the embedded clause is raised to the main clause; that is, there is a raising from subject to object. In terms of exceptional case assignment, it is proposed that the predicate of the main clause can assign case to the subject of the subordinate clause exceptionally. However, how come an argument that is controlled and assigned role by the predicate of the main clause becomes the argument of another verb and thus the subject of the subordinate clause? How does the subject of a sentence in Turkish realise in accusative case? Although sanmak, the predicate of the main clause, is a two-valence verb, how is it able to assign case to three arguments? Accepting that there is an exceptional case assignment does not comply with the valency of verbs, case criterion or object criterion. Raising is an important starting point for the analysis of doubleobject-like constructions; however, it does not sufficiently explain these structures. It is seen that in the formation of these structures syntactic noun incorporation and verb modifier processes are used. It can be asserted that the subject of the embedded clause that is raised to the object of the main clause first gets into a syntactic noun incorporation with the predicate of the main clause, and then it transforms into a syntactic construct that can be defined as verb modifier. It can also be asserted that the construct that is the predicate of the subordinate clause and behaves like an object is exposed to a syntactic and semantic corporation with the verb no matter what case it is assigned and what morphological structure it has. When the subject of the embedded clause is raised to the object of the main clause, the embedded clause becomes small clause. Noun incorporation goes through another syntactic and semantic formation. When the subject of the small clause is raised to object, the verbal feature continues the noun incorporation relationship as well as being exposed to a new process where it is assigned a new syntactic and semantic function. That is why, the relationship between the verbal feature of the small clause and predicate of the main clause cannot be explained only by noun incorporation. There is a different syntactic and semantic relationship between the predicate of the small clause and the predicate of the main clause. While the subject of the small clause realises as the ogical argument of the predicate of the main clause when it is raised to object, the predicate of the small clause functions as a modifier on the surface structure, not as a logical argument of the verb. Although the predicate of the small clause and verb realise as individual syntactic constructs, together they define an action, a process. The predicate of the main clause and the predicate of the small clause identify the linguistic function of the construct that is raised to object and assign its role. The process of the predicate of the small clause and the predicate of the main clause determining the syntactic arguments and linguistic functions of these arguments, and them assigning roles to these arguments give them an incorporated verb predicate feature (Rapport, 1995: 168, 169). The constructs that are generally regarded as double-object constructions in the literature on Turkish are actually structures consisting of a small clause transformed from embedded clauses. In the structures where embedded clauses transform into small clauses, it is seen that the subject of the embedded clauses is raised to the object-1 of the main clause while the verbal feature behaves like object-2 although it might be interpreted as indirect object or adverbial phrase. However, the predicate of the small clause is exposed to a new syntactic process with the predicate of the main clause on the surface structure. In the syntactic process, the predicate of the small clause has undertaken a verb modifier function. Verb modifier has a substantially individual syntactic and semantic function that differs from grammatical functions such as subject, object or indirect object which verb modifier cannot be defined with. Verb modifier is a significant syntactic feature that does not realise within the valence of the verb but is one of the features that are bound to verb on the surface structure, completes the verb semantically and is able to assign roles to the other arguments together with the verb. In syntactic analyses of Turkish, it is necessary to differentiate and identify the verb modifier whose syntactic location cannot be changed and which is adjacent to the verb. It is observed that verb modifier in Turkish generally occurs with cognitive verbs such as sanmak, zannetmek, bilmek, and in constructs where intransitive verbs formed with noun+olmak are made causative