Hi @Grant, glad to hear you are interested. Off the top of my head is Berwick and Chomsky's 2016 book [_Why Only Us_](https://mitpress.mit.edu/books/why-only-us) which is a fairly recent (and not too technical) summary of the main results of the field in the past few decades (the first chapter in [Carnie et al.'s handbook](https://books.google.co.uk/books?id=K091AwAAQBAJ&source=gbs_book_other_versions) is also a nice overview). For example, on page 8 Berwick and Chomsky say (emphases mine):

> [Human language syntactic structure] has at least three key properties, all captured by minimalist system assumptions: (1) human language syntax is **hierarchical**, and is **blind to considerations of linear order**, with linear ordering constraints reserved for **externalization**; (2) the particular hierarchical structures associated with sentences affects their interpretation; and (3) there is no upper bound on the depth of relevant hierarchical structure. Note that if all this is true, then observation (1) implies that any adequate linguistic theory must have some way to construct arrays of hierarchically structured expressions, while **ignoring linear order**; while (2) implies that structure (in part) fixes interpretation at the level of “meaning.” Finally, (3) implies that these expressions are potentially infinite. These then are the minimal properties any adequate syntactic theory must encompass and that’s why they are part of the minimalist account.

So it is more or less a consensus (or at least a prevalent idea) among theoretical linguists nowadays that the structure-building module (i.e. syntax) and the structure-externalizing module (i.e. phonology) should be studied and formalized separately, where only the former is relevant to the structure-interpreting module (i.e. semantics). This perspective makes the general architecture of the human language faculty into a so-called "Y-model"; see e.g. [this article](https://ddd.uab.cat/pub/cjol/16956885v8/16956885v8p141.pdf) for a themed discussion. Hope this helps. :-)

> [Human language syntactic structure] has at least three key properties, all captured by minimalist system assumptions: (1) human language syntax is **hierarchical**, and is **blind to considerations of linear order**, with linear ordering constraints reserved for **externalization**; (2) the particular hierarchical structures associated with sentences affects their interpretation; and (3) there is no upper bound on the depth of relevant hierarchical structure. Note that if all this is true, then observation (1) implies that any adequate linguistic theory must have some way to construct arrays of hierarchically structured expressions, while **ignoring linear order**; while (2) implies that structure (in part) fixes interpretation at the level of “meaning.” Finally, (3) implies that these expressions are potentially infinite. These then are the minimal properties any adequate syntactic theory must encompass and that’s why they are part of the minimalist account.

So it is more or less a consensus (or at least a prevalent idea) among theoretical linguists nowadays that the structure-building module (i.e. syntax) and the structure-externalizing module (i.e. phonology) should be studied and formalized separately, where only the former is relevant to the structure-interpreting module (i.e. semantics). This perspective makes the general architecture of the human language faculty into a so-called "Y-model"; see e.g. [this article](https://ddd.uab.cat/pub/cjol/16956885v8/16956885v8p141.pdf) for a themed discussion. Hope this helps. :-)