The dynamics of prepositions and prepositional constructions in Spanish

该网络旨在调查西班牙语介词结构以及介词成分在其中的作用，例如，Distinto a/de‘Different to’，vehículo de/a Motor‘Motor Vehicle’或Contactar con/a alguien‘Contact sb’，这在罗曼语中是很典型的。介词和介词结构都是极具异质性和争议性的范畴，在基于范畴的语言学方法以及关于语言习得和加工的心理和神经语言学研究中都没有统一的分类。这在很大程度上是因为介词和介词结构都处于词汇-语法连续体上，并且包括传统上被认为是语法化和词汇化的元素。为了掌握介词和介词成分的特殊地位，我们采用了基于用法的构式语法方法，将介词视为一个新兴类别，它是从构式中的原型槽学习的，需要不断地重组，原则上不会达到静态终点。在建构主义方法中，我们假设从完全图式结构(如[N Prep N])到完全词汇填充结构(如[vehículo de Motor])，从完全图式结构(如[N Prep N])到完全词汇填充结构(如[N de N])。网络项目的基本假设是，不同图式水平之间和内部的相互作用和可变性决定了介词的性质和作用。我们认为，这种视角的改变至少在三个方面有助于更好地理解介词的性质。首先，我们不研究孤立的项目，而是关注更大的语言单位(或结构)以及由它们连接的各自(词汇、语法或话语)元素。其次，词汇和句法之间的连续体的假设解决了词汇和语法之间不明确的界限，无论是对于介词还是对于它们所在的结构。第三，我们的观点强调了具体符号和更多图式模式之间的动态关系，这些模式处于不同的复杂和抽象水平。依靠共时和历时语料库分析，可以在组合和聚合联想网络中捕捉到这些信息。为了确定在构式网络中可以被认为是相关的构式在个别说话者中是否也存在认知联系，我们依靠心理语言学的方法，如启动和眼球跟踪。在这一点上，我们批判性地考察了中级(部分填充)结构的作用，其中介词‘支柱’被认为是介词结构处理的中心角色。我们的发现及其对西班牙语的理论启示也可以扩展到跨语言比较的角度。

The network aims at investigating Spanish prepositional constructions and the role of the prepositional elements within them, e.g. distinto a/de ‘different from’, vehículo de/a motor ‘motor vehicle’ or contactar con/a alguien ‘contact sb’, which are typical for the Romance languages. Both prepositions and prepositional constructions are extremely heterogeneous and disputed categories and defy a uniform classification in category-based linguistic approaches as well as in psycho- and neurolinguistic studies on language acquisition and processing. This is mostly due to the fact that prepositions and prepositional constructions both lie on a lexicon-grammar continuum and include elements that have traditionally been considered grammatical(ized) and lexical(ized) items.In order to grasp the particular status of prepositions and prepositional elements, our network adopts a usage-based Construction Grammar approach, in which prepositions are seen as an emergent category that is learned from prototypical slots in constructions, which are subject to constant reorganization and where a static endpoint is in principle never attained. In constructionist approaches, we assume different levels of schematicity going from fully schematic constructions, such as [N Prep N], over partially filled constructions, such as [N de N], to fully lexically filled constructions, such as [vehículo de motor]. The underlying hypothesis for the network projects is that the interaction and variability between and within different levels of schematicity determine both the nature and the role of prepositions. We assume that this change of perspective contributes to a refined understanding of the nature of prepositions in at least three ways. First, we do not investigate isolated items but bring into focus larger linguistic units (or ‘constructions’) as well as the respective (lexical, grammatical or discursive) elements that are connected by them. Second, the assumption of a continuum between lexicon and syntax resolves the unclear lexical vs. grammatical cut-off line, both for prepositions and for the constructions in which they occur. Third, our perspective highlights the dynamic relationship between concrete tokens and more schematic patterns at various levels of complexity and abstraction. Relying on synchronic and diachronic corpus analyses, these can be captured in syntagmatic and paradigmatic associative networks. In order to determine whether constructions that can be assumed to be related in a constructional network are also cognitively linked in individual speakers, we rely on psycholinguistic methods such as priming and eye-tracking. In this vein, we critically examine the role of mid-level (partially filled) constructions, in which prepositional ‘pillars’ are attributed a central role for the processing of prepositional constructions. Our findings and their theoretical implications for Spanish can also be extended towards a cross-linguistic comparative perspective.