37.1248, Reviews: Towards a label-less grammar: Joanna Wall (2025)

[The LINGUIST List]

LINGUIST List: Vol-37-1248. Sun Mar 29 2026. ISSN: 1069 - 4875.

Subject: 37.1248, Reviews: Towards a label-less grammar: Joanna Wall (2025)

Moderator: Steven Moran (linguist at linguistlist.org)
Managing Editor: Valeriia Vyshnevetska
Team: Helen Aristar-Dry, Mara Baccaro, Daniel Swanson
Jobs: jobs at linguistlist.org | Conferences: callconf at linguistlist.org | Pubs: pubs at linguistlist.org

Homepage: http://linguistlist.org

Editor for this issue: Helen Aristar-Dry <hdry at linguistlist.org>

================================================================


Date: 29-Mar-2026
From: Philip Pellino [pellinop at gmail.com]
Subject: Linguistic Theories: Joanna Wall (2025)


Book announced at https://linguistlist.org/issues/36-1043

Title: Towards a label-less grammar
Series Title: LOT Dissertation Series
Publication Year: 2025

Publisher: Netherlands Graduate School of Linguistics / Landelijke
(LOT)
           http://www.lotpublications.nl/
Book URL: https://dx.medra.org/10.48273/LOT0686

Author(s): Joanna Wall

Reviewer: Philip Pellino

SUMMARY
Joanna Wall’s (2024) dissertation is assuredly a contribution to
Minimalist theorizing and forges ahead in the true spirit of how that
program was envisioned. That is, it fulfills the goal handed to us by
Chomsky (2021), i.e., to always be sharpening the Strong Minimalist
Thesis (SMT), and she undertakes this by innovating and pointing us
“Towards a Label-less Grammar”. Having said that, if you are a person
who is inclined to believe that the success of a research work in
linguistics can be measured by the extent to which it forces you to
reassess fundamentals, and you’ve been hunting for some fresh puzzles,
then this dissertation is both a success, and an opportunity for you.
Despite the title, I believe that this work is probably of interest to
more linguists than one might at first be inclined to think. By this I
mean that, generally speaking, debates on minimal computation are not
everyone’s cup of tea. However, while on the surface this dissertation
appears to be joining a very particular conversation about theoretical
syntactic machinery, there are actually quite a few architectural and
conceptual matters to ponder. Overall, I feel that the empirical
coverage that is attempted does allow a wider audience of
syntacticians, morphologists and semanticists to engage with it.
Because the author spends a good deal of time showing how the system
will tackle a variety of fundamental grammatical issues—and because
breadth and depth are always a bit at odds—there must be a tiny
sacrifice when it comes to reviewing previous studies and/or
re-presenting the argumentation surrounding the background material.
This means that the more advanced you are in your studies and
research, the more you will enjoy this book. Moreover, this book is
about big picture ideas. Therefore, it is best appreciated by
researchers who have spent time thinking about what the properties and
architecture of the Human Faculty of Language might or should be and,
thus, are familiar with previous debates surrounding those topics. So,
it's not necessarily the kind of book that you might point a first
year graduate student toward, but would be a nice work to build a
graduate seminar around. With that in mind, the large scope of
“Towards a Label-less Grammar” may be considered unavoidable, but a
positive quality; those of us seeking something to write about can
easily snatch up this dissertation, and get to work seeing if we can
debunk Wall’s system as we test it against our own favorite phenomena.
With this background in place, I’d like to summarize the
dissertation’s key ideas.
The primary innovation is obvious from the title: to lay out a
framework for syntactic structure building that operates free of any
category (sub)labels whatsoever. Wall’s starting position is simply
the Minimalist ideal which seeks to uncover what the “minimal design
specifications” are for the computational system that is Human
Language. Wall argues that this requires acknowledging that category
labels are artificial elements and, therefore, antithetical to the
Strong Minimalist Thesis (SMT) as outlined in Chomsky (2000). That is,
these labels largely violate both the Interpretability Condition and
the Inclusiveness Condition. Recall that the Interpretability
Condition states that the only allowable syntactic features are those
which are absolutely required for the Sensory Motor (SM) and
Conceptual Intentional (CI) interfaces to make use of the output of
Narrow syntax. In tandem, the Inclusiveness Condition stipulates that
features can not be introduced by Narrow syntax. Attempting to truly
satisfy these conditions, Wall dispenses with the following
(sub)labels: lexical category labels, functional category labels,
X-bar theoretic sub-labels and uninterpretable features. In sum, this
leaves a system which is relatively unlike those which many of us are
used to thinking about.
Presumably, linguists imagined that these (sub)labels were necessary,
and were doing some kind of useful computational work. And so,
obviously, Wall must redistribute the computational labor that they
were responsible for, as well as reimagine how that computational
labor is executed. In order to do this, Wall builds off a variety of
literature advancing the machinery in this direction, most notably:
Boeckx (2014), Borer (2005a, 2005b, 2013), Harbour (2014), Kayne
(2013, 2023) and Zwart (2004, 2012). What Wall delivers is an
extension of the work begun by these researchers, to which a variety
of innovations are contributed. The dissertation advances its
argumentation with the following structure: Chapters 1 and 2 provide
an overview of the general ideas behind the pursuit for a Label-less
Grammar; Chapters 3, 4 and 5 take up issues related to the syntactic
objects traditionally referred to as Nouns  (plurality, mass-count,
etc);  Chapter 6 examines Definiteness and Proper Names; Chapter 7
begins the discussion surrounding syntactic objects traditionally
referred to as Verbs; Chapters 8, 9, and 10 analyze Direct Objects,
Telicity and Aksionsart; Chapter 11 tackles well-known Present Tense
quandaries; and Chapter 12 concludes the work with a summary and some
speculation on required future research. Each chapter does set some
space aside for a brief comparison of the ways that the given
phenomena are handled by other Generative approaches, but many
observations are simply discussed along the way, as matters are
reanalyzed, nuanced and/or built upon.
In what follows, I will attempt to lay out very parsimoniously what
Wall’s innovations are, and how they can be employed to build
(morpho-)syntactic structures. To begin, I will explain how the
Lexicon is envisioned to function, and subsequently turn to Wall’s
conceptualization of Merge.
As you begin to imagine what Wall’s label-less grammatical system
looks like, it is important to keep in the forefront of your mind the
fact that it lacks some traditional items like nouns or verbs. What
Wall assumes is a single pre-syntactic lexicon very similar to Borer’s
encyclopedia (2005a, b), and similarly refers to the items contained
in it as ‘listemes.’ Each listeme entry is (potentially) tripartite
and allows specifications for: Phonological content, Conceptual
content, and an Entity denotation. The familiar lexical and functional
category distinctions are the result of the presence, absence, and/or
subsequent combinations of these three content specifications in
conjunction with the Merger of multiple other listemes. For example, a
determiner-type element like /the/ will have Phonological content,
Conceptual content and an Entity denotation (a finite singleton set).
While an indefinite-type element like /a/ will have Phonological
content, and an Entity denotation (finite singleton set), but no
Conceptual content. Further, if we take a listeme for something that
we might traditionally think of as a noun or a verb such as /book/, it
will have Phonological content, Conceptual content but no Entity
denotation. Now, while the Phonological content and Conceptual content
(World knowledge-type information) is relatively self-explanatory, the
idea of an Entity denotation is potentially not. Simply put,
‘entities’ are proposed to be ontological primitives, and a listeme
can be specified for either a) a finite, singleton set — referring to
a singularity of objects or b) an infinite, non-singleton set —
referring to a plurality of objects. To reiterate, it is the
interaction of the specifications of these syntactic objects that give
rise to possible and impossible combinations of them and hence,
possible and impossible complex structures. As the reader probably
recalls, according to Chomsky (2004), it is the C-I interface that we
want to be responsible for “filtering” licit from illicit structures.
And so now, we’ll turn to the way in which Wall argues that this
occurs.
To handle the label-less grammar proposed here, Wall introduces a new
Conceptual Interface (C-I) condition called ‘Merge as Specify’ (MaS).
MaS is evaluative of a species of Pair-Merge that creates an
asymmetric relationship between two objects, A and B. Basically, when
A is merged with B, A forms a semantic part-of relationship with B. Of
course, this is only possible if another condition is met: the
semantic ‘extent’ of B must be greater than that of A. Here, we want
to understand a syntactic object’s extent as the semantic breadth of
its Conceptual content. And, if it has an Entity denotation, then the
particular set of entities denoted there. So, A specifies B insofar as
it narrows the semantic properties expressed by B in its selection of
some part-of them. The asymmetry results as the semantic properties of
B are more foundational than those of A. In this sense, the condition
underlying MaS is natural and familiar. It is akin to what is attested
in cases of compounding and adjectival modification. Allow me to try
and construct a simple example. Let’s consider a structure like ‘green
pond’. Here, the semantics of /pond/ are further specified insofar as
at least one property in the set of properties that denote /pond/,
that of Color, is specified as /green/. In the case of MaS, we have a
corollary for complex syntactic objects like ‘The boy eats.’ Here, the
semantics of ‘eats’ (our syntactic object B — the Mergee) are further
specified through a part-of relationship established by Merging ‘The
boy’ (our syntactic object A — the Merger). Some properties in the set
of properties that constitute what it means ‘to eat’ are picked out.
Namely, those that can be specified by ‘The boy’. Stated a bit
differently, we narrow the semantic space (or extent) of ‘eats’ in
terms of the extent of ‘The boy.’ That is, we end up with the part of
the extent of ‘eats’ that can be overlaid by the semantic extent of
‘The boy.’ In sum, our new object AB is conceptually the mapping of A
onto B.
Now, Wall does need to introduce a stipulation—the Specificity
Hierarchy—in order to help manage Merge under MaS. And, this hierarchy
is exactly what it sounds like: a hierarchy of the specificity of
syntactic objects relative to one another and that is modeled upon the
tripartite division of the specifications that comprise the listemes.
Thus, according to the Specificity Hierarchy, a syntactic object with
a specification(s) for a finite singleton set of entities (+
conceptual content), is-more-specific-than one specified for
Conceptual content (+infinite non-singleton sets of entities),
is-more-specific-than one which has only a specification for an
infinite non-singleton set of entities. Tying the Specificity
Hierarchy and the MaS condition together, what this means is that a
syntactic object that is higher in the hierarchy (more specific) must
Merge with something lower in the hierarchy (less specific). This is
fairly straightforward: a more complex object (a set of entities) can
not form a part-of relationship with something that is less complex
than itself (a single entity); and an object with Conceptual content
can not be part-of an object that lacks Conceptual content. So for
example, the Merger of two items like /a/ and /the/ would be blocked,
but a Merger of /the/ and /book/ is acceptable. This is because /a/
and /the/ both have an Entity denotation of finite singleton set, and
therefore, are on the same level of the hierarchy, and can not combine
with one another. Alternatively, /book/ simply has Conceptual content,
and being lower on the hierarchy is free to combine with either /a/ or
/the/.
Now, an interesting consequence of MaS and the Specificity Hierarchy
concerns the tree structures that it leads to. Because licit Merge is
driven by the satisfaction of the Specificity Hierarchy in regard to
the combinations of lexical specifications, the system permits
derivations, and ultimately, constituencies that are in general barred
in other generative models. So, standardly, we think of an expression
with multiple underlying trees as coinciding with instances of
ambiguity. That is, the ambiguity is explained as the result of two
(or more) possible trees. However, in the system constrained by MaS,
it is actually possible to end up with multiple trees that correspond
to a single unique meaning. Moreover, there is no real sense in which
derivations are “bottom-up” or “top-down”. This possibility for
multiple trees results in cases when the listemes are such that they
can combine with one another in more than one sequence to form the
same linear order. (And this system assumes that Linear Order is Merge
Order.) So, imagine something like ‘The bear eats berries.’ MaS allows
us to build this as either ‘The bear’ [DP] Merges with ‘eats berries’
[V DP ] to yield [DP [V DP]] or ‘The bear eats’ [DP V] Merges with
‘berries’ [DP] to yield [[DP V] DP]. Wall suggests that data from
coordination and right-node-raising potentially provide evidence for
the validity of these  alternate structures. In both of these kinds of
constructions, we find missing material which has received a variety
of analyses: it is displaced, it is deleted, there is multidominance,
etc. Instead, Wall hypothesizes that instead of such processes, it
might be the case that such constructions are simply composed of the
exact structures that MaS predicts are possible. Therefore, in the
case of something like ‘John stole and Bill ate a cake’, what we
really have is coordination of two [Subj V] sets, and subsequent Merge
of a single object: [[[Subj V]  [and [Subj V]]] a cake].
EVALUATION
For researchers who have spent a great deal of their time working on
specific complex construction types, especially those involving A-bar
phenomena (Internal Merge), then this dissertation really is an
invitation to join in a research agenda. That is, Wall doesn’t devote
any significant space to these kinds of issues. Instead, Wall must
focus on the fundamentals mentioned earlier in the overview of the
chapters (an impressive amount that there simply is not enough space
to discuss). However, in regard to Internal Merge, Wall’s hypothesis
is that its role may be significantly marginalized in a label-less
grammar, and speculates that it could potentially be abandoned.
Currently, she suggests that Internal Merge plays a role only in cases
of the “derived subjects of passives, unaccusatives and
anticausatives, and elements involved in A-bar movement
constructions." Obviously, a firmer stance can be taken only after a
significant amount of future work unfolds. But to reiterate, the
groundwork laid here is substantial, and so it is the potential
starting point for a very large number of subsequent projects.
As a final note, I think that researchers both in Linguistics and
other adjacent areas of Cognitive Science will find the psychological
underpinning of Merge as Specify to be intriguing. The idea that
forming coherent part-of relations (and hence, coherent Concepts) is
not what Syntax does but what Syntax is, is an extreme position on one
side of an ongoing and controversial debate regarding the connection
between Language and Thought. Although Wall does not make this an
explicit part of her discussion, I spent a lot of time thinking about
her proposal in terms of the potential contacts that it makes with
this literature, and I suspect that others will as well (See: Collins
2019, 2021; Dupre 2021; Fedorenko et al 2024; for some alternative
viewpoints on this issue). This is another positive attribute of
“Towards a Label-less Grammar”, and an additional example of its
success as a contribution to the Biolinguistic Program. Throughout, we
are forced to confront our assumptions about the Computational system
of Human Language for the purpose of achieving “genuine explanations”,
and to wrestle considerably with questions of what can be “purged”,
and what is indispensable (Chomsky 2021).
References:
Boeckx, Cedric. 2014. Elementary syntactic structures: Prospects of a
feature-free syntax. Cambridge: Cambridge University Press.
Borer, Hagit. 2005a. Structuring sense vol 1: In name only. Oxford:
Oxford University Press.
Borer, Hagit. 2005b. Structuring sense vol 2: The normal course of
events. Oxford: Oxford University Press.
Borer, Hagit. 2013. Structuring sense vol 3: Taking form. Oxford:
Oxford University Press.
Chomsky, Noam. 2000. “Minimalist Inquiries: The framework.” In: Roger
Martin, David Michaels, Juan Uriagereka, and Samuel J. Keyser (Eds.)
Step by Step: Essays on minimalist syntax in honor of Howard Lasnik
(pp 89 - 155). Cambridge, MA: The MIT Press.
Chomsky, Noam. 2004. “Beyond explanatory adequacy.” In: Adriana
Belletti (Ed.), Structures and Beyond: The cartography of syntactic
structures. Volume 3 (pp. 104 - 131). Oxford: Oxford University Press.
Chomsky, Noam. 2021. Chomsky, Noam. "Minimalism: Where are we now, and
where can we hope to go." Gengo Kenkyu (Journal of the Linguistic
Society of Japan) Vol 160 (pp 1 - 41).
Collins, Chris. 2019. Thought as Syntax.
https://ordinaryworkinggrammarian.blogspot.com/2019/10/thought-as-syntax.html.
Thursday, Oct 24, 2019 Ordinary Working Grammarian Blog.
Collins, Chris. 2021. A Conversation with Noam Chomsky about Formal
Semantics.
https://ordinaryworkinggrammarian.blogspot.com/2021/06/a-conversation-with-noam-chomsky-about.html.
Ordinary Working Grammarian Blog.
Dupre, Gabe. 2021. “What would it mean for Natural Language to be the
Language of Thought”. Linguistics and Philosophy 44 (pp. 773–812).
https://doi.org/10.1007/s10988-020-09304-9.
Fedorenko, E; S.T Piantadosi; E.A.F Gibson. 2024. “Language is
primarily a tool for communication rather than thought”. Nature 630
(pp. 575-586).
Harbour, David. 2014. “Paucity, abundance, and the theory of number.”
Language 90(1) (pp. 185 - 229).
Kayne, Richard. 2013. “Why are there no directionality parameters.”
Studies in Chinese Linguistics 34(1) (pp. 3 - 37).
Kayne, Richard. 2023. “Temporal/Linear order, antisymmetry, and
externalization.” RGG 45(2) (pp. 1 - 22).
Zwart, Jan-Wouter. 2004. Unary Merge. Handout from the paper presented
at the Tilburg University staff seminar Grammaticamodellen, Oct 28,
2004. https://www.let.rug.nl/~zwart/docs/ho04tilb.pdf
Zwart, Jan-Wouter. 2012. “Structure and Order: Asymmetric Merge.” In:
Cedric Boeckx (Ed.) The Oxford handbook of linguistic minimalism. (pp.
96 - 118). Oxford/New York, NY: Oxford University Press
About the Reviewer:
Philip Pellino holds a PhD in Linguistics from Michigan State
University, and specializes in the interface of Semantics and
Pragmatics with Syntax. His primary interests are: Negation,
Information Structure, Inferential Meaning, and the Philosophy of
Language. Professionally, he works as a Computational linguist.



------------------------------------------------------------------------------

********************** LINGUIST List Support ***********************
Please consider donating to the Linguist List, a U.S. 501(c)(3) not for profit organization:

https://www.paypal.com/donate/?hosted_button_id=87C2AXTVC4PP8

LINGUIST List is supported by the following publishers:

Bloomsbury Publishing http://www.bloomsbury.com/uk/

Cambridge University Press http://www.cambridge.org/linguistics

Cascadilla Press http://www.cascadilla.com/

De Gruyter Brill https://www.degruyterbrill.com/?changeLang=en

Edinburgh University Press http://www.edinburghuniversitypress.com

European Language Resources Association (ELRA) http://www.elra.info

John Benjamins http://www.benjamins.com/

Language Science Press http://langsci-press.org

Lincom GmbH https://lincom-shop.eu/

MIT Press http://mitpress.mit.edu/

Multilingual Matters http://www.multilingual-matters.com/

Narr Francke Attempto Verlag GmbH + Co. KG http://www.narr.de/

Netherlands Graduate School of Linguistics / Landelijke (LOT) http://www.lotpublications.nl/

Peter Lang AG http://www.peterlang.com

SIL International Publications http://www.sil.org/resources/publications


----------------------------------------------------------
LINGUIST List: Vol-37-1248
----------------------------------------------------------