In Goldbergian Construction Grammar (CxG), phrasal verb-independent Argument Structure Constructions (ASCs) take center stage. These ASCs account for novel uses of verbs and CxG thereby elegantly dispenses with implausible verb senses, and even keeps the number of verb senses to a minimum, often positing single lexical entries. This view of language representation emphasizes the importance of generalizations, while item-specific knowledge is largely downplayed. It has however been shown that such generalizations (1) fail to predict the difference in syntactic patterning of semantically related verbs (e.g. Iwata, 2008), and (2) cannot always account for the intended semantic interpretation (e.g. Boas, 2008). Boas (2003) therefore posits mini-constructions, which are lexically bound verb-specific constructions that account for the idiosyncratic syntactic behavior and semantic interpretation of verbs. To establish such mini-constructions, Boas calls for a fine-grained bottom-up identification of all the argument realization patterns of a verb according to the verb’s different senses. Such an approach would “[result] in a much more detailed lexicon in which verbs are associated with a number of conventionalized senses, each of which forms its own mini-construction that is a pairing of a form with a meaning” (2003: 37). Against this backdrop, this study takes a step forward in answering this call by adopting a corpus-based bottom-up approach to the formal and semantic patterning of the high-frequency verb make. High-frequency verbs represent an interesting case as on the one hand they are considered pathbreaking verbs for particular ASCs (Goldberg, 2006), while on the other they enter a wide range of syntactic configurations. This study analyzes all 1,975 occurrences of make in the one million-word British component of the International Corpus of English. In a first step, each occurrence is assigned a broad ASC. In a second step, the patterning of each occurrence is analyzed from a lexical verb-based perspective, largely drawing on Hanks’ lexically based methodology of Corpus Pattern Analysis (Hanks, 2013): form is annotated by assigning a valency structure to each instance, while meaning is annotated by assigning a semantic value to each valency-slot and an overall meaning to the pattern, thereby arriving at verb-specific form-meaning mappings, i.e. mini-constructions, bottom-up. Subsequently, it is possible to flesh out each ASC for mini-constructions that have narrower and verb-specific meanings, but that still inherit features from the ASC. Results suggest that (1) from a verb-based perspective, rather than being reduced to single lexical entries, verbs display rich patterns of polysemy that are needed to account for their syntactic patterning and (2) from a construction-based perspective, ASCs have lexically-bound patterns of polysemy with narrower meanings. This study hereby hopes to make a case for lexically-bound mini-constructions and, by extension, for the recognition of the importance of item-specificity alongside verb-independent generalizations. References Boas, H. C. (2003). Towards a lexical-constructional account of the locative alternation. In Carmichael, L., C.-H. Huang, and V. Samiian (eds.), Proceedings of the 2001 Western Conference in Linguistics vol. 13, 27-42. Boas, H. C. (2008). Resolving form-meaning discrepancies in Construction Grammar. In Leino, J. (ed.), Constructional reorganization. Amsterdam/Philadelphia: John Benjamins, 11-36. Goldberg, A. E. (2006). Constructions at work. Oxford: Oxford University Press. Hanks, P. (2013). Lexical analysis: norms and exploitations. Cambridge MIT press. Iwata, S. (2008). Locative alternation – a lexical-constructional approach. Amsterdam/Philadelphia: John Benjamins.