As algorithms have become ubiquitous in consequential domains, societal concerns about the potential for discriminatory outcomes have prompted urgent calls to address algorithmic bias. In response, a rich literature across computer science, law, and ethics is rapidly proliferating to advance approaches to designing fair algorithms. Yet computer scientists, legal scholars, and ethicists are often not speaking the same language when using the term ‘bias.’ Debates concerning whether society can or should tackle the problem of algorithmic bias are hampered by conflations (...) of various understandings of bias, ranging from neutral deviations from a standard to morally problematic instances of injustice due to prejudice, discrimination, and disparate treatment. This terminological confusion impedes efforts to address clear cases of discrimination. -/- In this paper, we examine the promises and challenges of different approaches to disambiguating bias and designing for justice. While both approaches aid in understanding and addressing clear algorithmic harms, we argue that they also risk being leveraged in ways that ultimately deflect accountability from those building and deploying these systems. Applying this analysis to recent examples of generative AI, our argument highlights unseen dangers in current methods of evaluating algorithmic bias and points to ways to redirect approaches to addressing bias in generative AI at its early stages in ways that can more robustly meet the demands of justice. (shrink)

ABSTRACT This article distinguishes nine senses of polarization and provides formal measures for each one to refine the methodology used to describe polarization in distributions of attitudes. Each distinct concept is explained through a definition, formal measures, examples, and references. We then apply these measures to GSS data regarding political views, opinions on abortion, and religiosity—topics described as revealing social polarization. Previous breakdowns of polarization include domain-specific assumptions and focus on a subset of the distribution’s features. This has conflated multiple, (...) independent features of attitude distributions. The current work aims to extract the distinct senses of polarization and demonstrate that by becoming clearer on these distinctions we can better focus our efforts on substantive issues in social phenomena. (shrink)

It is often claimed that the intentions of physicians are multiple, ambiguous, and uncertain—at least with respect to end-of-life care. This claim provides support for the conclusion that the principle of double effect is of little or no value as a guide to end-of-life pain management. This paper critically discusses this claim. It argues that proponents of the claim fail to distinguish two different senses of “intention,” and that, as a result, they are led to exaggerate the extent to which (...) clinical intentions in end-of-life contexts are ambiguous and uncertain. It argues further that physicians, like others who make life and death decisions, have a duty to get clear on what their intentions are. Finally, it argues that even if the principle of double effect should be rejected, clinical intentions remain ethically significant because they condition the meaning of extraordinary clinical interventions, such as that of palliative sedation. (shrink)

Recent findings in neuroscience strongly suggest that an object’s features (e.g., its color, texture, shape, etc.) are represented in separate areas of the visual cortex. Although represented in separate neuronal areas, somehow the feature representations are brought together as a single, unified object of visual consciousness. This raises a question of binding: how do neural activities in separate areas of the visual cortex function to produce a feature-unified object of visual consciousness? Several prominent neuroscientists have adopted neural synchrony and attention-based (...) approaches to explain object feature binding. I argue that although neural synchrony and attentional mechanisms might function to disambiguate an object’s features, it is difficult to see how either of these mechanisms could fully explain the unity of an object’s features at the level of visual consciousness. After presenting a detailed critique of neural synchrony and attention-based approaches to object feature binding, I propose interactive hierarchical structuralism (IHS). This view suggests that a unified percept (i.e., a feature-unified object of visual consciousness) is not reducible to the activity of any cognitive capacity or to any localized neural area, but emerges out of the interaction of visual information organized by spatial structuring capacities correlated with lower, higher, and intermediate levels of the visual hierarchy. After clarifying different notions of emergence and elaborating evidence for IHS, I discuss how IHS can be tested through transcranial magnetic stimulation and backward masking. In the final section I present some further implications and advantages of IHS. (shrink)

Drawing on convergent work in a broad range of disciplines, this article uses the tipping point paradigm to frame a new account of how early human ancestors may have first broken free from, as Bickerton calls it, the “prison of animal communication.” Under building pressure for an enhanced signaling system capable of supporting joint attentional-intentional activities, a cultural tradition of disambiguated indexical pointing, combined with increasingly sophisticated mindreading circuitry and prosocial tendencies, may have sparked the first in the series of (...) biocultural explosions that led from a simple protolanguage to fully modern human language. This account successfully integrates at least ten other competing hypotheses, and is shown to pass nine important tests that have been proposed for language origin scenarios of its type. (shrink)

This essay presents an analysis of the conversational role of disambiguation, with special attention to disambiguating parentheticals such as 'bats, the furry animals, are not easy to find'. The essay proposes an enriched representation of conversational states as pairs of an interpretation function and standard common belief, it represents disambiguations within the ensuing framework, and, on the basis of these conceptual tools, it proposes a systematic picture of the updates achieved by disambiguating parentheticas.

This article uses Jeremy Bentham's notion of disambiguation, which links language to power and ‘sinister interest’, to analyse criticisms of the Royal Academy of Arts by Benthamites and Philosophic Radicals at the Select Committee on Arts and Manufactures of 1835/6. This practice of disambiguation aimed to produce a distinction between the Royal Academy of Arts and the publicly funded art school. I situate this activity within the linguistic turn taken by Bentham's ethics, and its relevance to a dilemma (...) of pedagogy in commercial society framed by Adam Smith. Smith's dilemma turns on the conflict between the requirement for a pedagogy that conforms to the principle of free trade, and an equally binding requirement for a virtue ethical model of pedagogy that offers a remedy for the corrupting effects of commerce on character. Adam Smith's support for private academies of art asserted a hierarchy of virtue ethics over utility, thus safeguarding autonomous ethical reasoning within capitalist forms of social life. Bentham's thought, in contrast, eschews the link between ethics and character, and places ethics itself within normative rules of language and cognition. (shrink)

Next SectionThis paper proposes an extension of Dependency Tree Semantics (DTS), an underspecified formalism originally proposed in Robaldo (2007). The crucial advantage of DTS as compared to other contemporary proposals is its ability to represent Independent Set (IS) readings (a.k.a. scopeless readings), e.g. cumulative and collective readings. DTS achieves the expressivity needed to represent IS readings because it underspecifies Skolem-like functional dependencies. This paper extends DTS by introducing additional meta-constraints in First-Order Logic dedicated to disambiguating underspecified structures. However, it is (...) worth noting that the meta-constraints are independent from DTS, and could be easily re-implemented in any underspecified theory. The meta-constraints achieve flexible disambiguation, in the sense that they allow ambiguous structures to be specified independently of the linguistic (rather than logical) knowledge used. Secondly, DTS, equipped with the meta-constraints, becomes an expressively complete formalism in the sense of Ebert (2005). Expressive completeness is a desirable property for an underspecified formalism, because it allows a representation for each possible subset of available readings. (shrink)

We bring together two ideas that are widely accepted in the literature: (i) conditionals with modal consequents are systematically ambiguous between a reading on which the if-clause restricts the modal, and a reading on which the if-clause introduces an epistemic supposition and the modal is unrestricted; (ii) ‘will’ is a modal. Together, (i) and (ii) imply that ‘will’-conditionals are systematically ambiguous between a reading as restricted historical claims, and one as unrestricted historical claims made under a supposition. We argue that (...) this prediction is correct and that the existence of the relevant ambiguity has far-reaching repercussions for several important debates concerning conditionals. (shrink)

This article focuses on key roles that the ill-defined concept of ‘public benefit’ plays in accessing the public health data held by the UK’s National Health Service. Using the concept of the ‘trade-off fallacy’, this article argues that current data access and governance structures, based on particular construals of public benefit in the context of public health data, largely negate the possibility of effective control by individuals over future uses of personal health data. This generates a health data version of (...) the trade-off fallacy that enables widespread involvement of commercial actors in personal data, despite public concerns over commercial involvement in, and potential exploitation of, public health data. The article suggests that, despite ostensibly robust regulatory and governance structures, this publicly held data is potentially subject to similar logics of accumulation as seen elsewhere in the digital economy, highlighting the inadequacies of current data regulatory frameworks in the digital era. (shrink)

Talk of mechanisms is ubiquitous in the natural sciences. Interdisciplinary fields such as biochemistry and pharmacy frequently discuss mechanisms with the assistance of diagrams. Such diagrams usually depict entities as structures or boxes and activities or interactions as arrows. While some of these arrows may indicate causal or componential relations, others may represent temporal or operational orders. Importantly, what kind of relation an arrow represents may not only vary with context but also be underdetermined by empirical data. In this manuscript, (...) we investigate how an analysis of pharmacological mechanisms in terms of producing and underlying mechanisms—as discussed in the contemporary philosophy of science—may shed light on these issues. Specifically, we shall argue that while pharmacokinetic mechanisms usually describe causal chains of production, pharmacodynamics tends to focus on mechanisms of action underlying the in vivo effects of a drug. Considering the action of thyroid gland hormones in the human body as a case study, we further demonstrate that pharmacodynamic schemes tend to incorporate entities and interactions on multiple levels. Yet, traditional pharmacodynamic schemes are sketched “flat”, i.e., non-hierarchically. We suggest that transforming flat pharmacodynamic schemes into mechanistic multi-level representations may assist in disentangling the different kinds of mechanisms and relations depicted by arrows in flat schemes. The resulting Baumkuchen model provides a powerful and practical alternative to traditional flat schemes, as it explicates the relevant mechanisms and relations more clearly. On a more general note, our discussion demonstrates how pharmacology and related disciplines may benefit from applying concepts from the new mechanist philosophy to guide the interpretation of scientific diagrams. (shrink)

Russo and Williamson claim that establishing causal claims requires mechanistic and difference-making evidence. In this article, I will argue that Russo and Williamson's formulation of their thesis is multiply ambiguous. I will make three distinctions: mechanistic evidence as type vs object of evidence; what mechanism or mechanisms we want evidence of; and how much evidence of a mechanism we require. I will feed these more precise meanings back into the Russo–Williamson thesis and argue that it is both true and false: (...) two weaker versions of the thesis are worth supporting, while the stronger versions are not. Further, my distinctions are of wider concern because they allow us to make more precise claims about what kinds of evidence are required in particular cases. (shrink)

Linking constructions involving dሇ (DE) are ubiquitous in Chinese, and can be translated into English in many different ways. This is a major source of machine translation error, even when syntaxsensitive translation models are used. This paper explores how getting more information about the syntactic, semantic, and discourse context of uses of dሇ (DE) can facilitate producing an appropriate English translation strategy. We describe a finergrained classification of dሇ (DE) constructions in Chinese NPs, construct a corpus of annotated examples, and (...) then train a log-linear classifier, which contains linguistically inspired features. We use the DE classifier to preprocess MT data by explicitly labeling dሇ (DE) constructions, as well as reordering phrases, and show that our approach provides significant BLEU point gains on MT02 (+1.24), MT03 (+0.88) and MT05 (+1.49) on a phrasedbased system. The improvement persists when a hierarchical reordering model is applied. (shrink)

In contrast to Borsboom who distinguishes between manifest and latent variables on epistemic grounds in terms of “epistemic accessibility,” I advocate a demarcation on pragmatic grounds. The latter way of understanding this distinction does justice to the intuitions driving Borsboom’s account, but avoids unnecessary epistemic complications. I then turn to two cases, the Flynn Effect and the case of psycho-educational assessment, and show an equivocal understanding of one latent variable, Spearman’s g, has led some researchers to draw paradoxical conclusions regarding (...) cognitive ability. (shrink)

Fregeans face the difficulty finding a notation for distinguishing statements about the sense or meaning of an expression as opposed to its reference or denotation. Famously, in "On Denoting", Russell rejected methods that begin with an expression designating its denotation, and then alter it with a "the meaning of" operator to designate the meaning. Such methods attempt an impossible "backward road" from denotation to meaning. Contemporary neo-Fregeans, however, have suggested that we can disambiguate _with_, rather than _against_, the grain, by (...) using a notation that begins with expressions designating senses or meanings, and then alters them with a "the denotation of" operator to designate the denotation. I show that in his manuscripts of 1903–05, Russell both considered and rejected a similar notation along with the metaphysical suppositions underlying it. This discussion sheds light on the evolution of Russell's thought, and may yet be instructive for ongoing debates. (shrink)

The resolution of ambiguities is one of the central problems for Machine Translation. In this paper we propose a knowledge-based approach to disambiguation which uses Description Logics (dl) as representation formalism. We present the process of anaphora resolution implemented in the Machine Translation systemfast and show how thedl systemback is used to support disambiguation.The disambiguation strategy uses factors representing syntactic, semantic, and conceptual constraints with different weights to choose the most adequate antecedent candidate. We show how these (...) factors can be declaratively represented as defaults inback. Disambiguation is then achieved by determining the interpretation that yields a qualitatively minimal number of exceptions to the defaults, and can thus be formalized as exception minimization. (shrink)

This article proposes a way to disambiguate the evaluative states currently identified as “adaptive preferences” in development literature. It provides a brief analysis of Serene Khader's Deliberative Perfectionist Approach, and demonstrates that distinguishing between adaptive states has important implications for the theory and practice of development intervention. Although I support Khader's general approach and consider my project to be complementary, I argue that the term preferences be replaced with four distinct terms: beliefs, choices, desires, and values. Distinguishing among adaptive states (...) can serve to prevent inappropriate intervention and appreciate the costs of transforming inappropriate adaptations. I argue that adaptive values are especially problematic, given how central a person's values are to their sense of meaning and self. Attempts to transform adaptive values are likely to produce internal conflict, resulting in psychological distress and diminished agency. Furthermore, some values preclude deliberation and comparison given their communal status as infinitely valuable. To deliberate about sacred values is to violate them. The emotional and psychological damage that may result from value transformation is thus likely to be extensive, and must be taken into account when determining whether, and what type of, intervention is justified. (shrink)

I distinguish between Old Contextualism, New Contextualism, and the Multiple Concepts Theory. I argue that Old Contextualism cannot handle the following three problems: (i) the disquotational paradox, (ii) upward pressure resistance, (iii) inability to avoid the acceptance of skeptical conclusions. New Contextualism, in contrast, can avoid these problems. However, since New Contextualism appears to be a semanticized mirror image of MCT, it remains unclear whether it is in fact a genuine version of contextualism.

Pessoa et al. (1998a) underexposed the broad and rich variety of stimuli in the amodal completion domain. The disambiguation of occluded shapes depends on very specific figural properties. Elaborations on such disambiguations of rich and complex stimuli, tied up with a visuo-cognitive origin of amodal completion, further position Pessoa et al.'s considerations on neural filling-in and the personal-subpersonal distinction.

The disambiguation of a syntactically ambiguous sentence in favor of a less preferred parse can lead to slower reading at the disambiguation point. This phenomenon, referred to as a garden‐path effect, has motivated models in which readers initially maintain only a subset of the possible parses of the sentence, and subsequently require time‐consuming reanalysis to reconstruct a discarded parse. A more recent proposal argues that the garden‐path effect can be reduced to surprisal arising in a fully parallel parser: (...) words consistent with the initially dispreferred but ultimately correct parse are simply less predictable than those consistent with the incorrect parse. Since predictability has pervasive effects in reading far beyond garden‐path sentences, this account, which dispenses with reanalysis mechanisms, is more parsimonious. Crucially, it predicts a linear effect of surprisal: the garden‐path effect is expected to be proportional to the difference in word surprisal between the ultimately correct and ultimately incorrect interpretations. To test this prediction, we used recurrent neural network language models to estimate word‐by‐word surprisal for three temporarily ambiguous constructions. We then estimated the slowdown attributed to each bit of surprisal from human self‐paced reading times, and used that quantity to predict syntactic disambiguation difficulty. Surprisal successfully predicted the existence of garden‐path effects, but drastically underpredicted their magnitude, and failed to predict their relative severity across constructions. We conclude that a full explanation of syntactic disambiguation difficulty may require recovery mechanisms beyond predictability. (shrink)

We describe a method for automatic word sense disambiguation using a text corpus and a machine- readable dictionary (MRD). The method is based on word similarity and context similarity measures. Words are considered similar if they appear in similar contexts; contexts are similar if they contain similar words. The circularity of this definition is resolved by an iterative, converging process, in which the system learns from the corpus a set of typical usages for each of the senses of the (...) polysemous word listed in the MRD. A new instance of a polysemous word is assigned the sense associated with the typical usage most similar to its context. Experiments show that this method performs well, and can learn even from very sparse training data. (shrink)

There is growing interest in the effect of sound on visual motion perception. One model involves the illusion created when two identical objects moving towards each other on a two-dimensional visual display can be seen to either bounce off or stream through each other. Previous studies show that the large bias normally seen toward the streaming percept can be modulated by the presentation of an auditory event at the moment of coincidence. However, no reports to date provide sufficient evidence to (...) indicate whether the sound bounce-inducing effect is due to a perceptual binding process or merely to an explicit inference resulting from the transient auditory stimulus resembling a physical collision of two objects. In the present study, we used a novel experimental design in which a subliminal sound was presented either 150 ms before, at, or 150 ms after the moment of coincidence of two disks moving towards each other. The results showed that there was an increased perception of bouncing when the subliminal sound was presented at or 150 ms after the moment of coincidence compared to when no sound was presented. These findings provide the first empirical demonstration that activation of the human auditory system without reaching consciousness affects the perception of an ambiguous visual motion display. (shrink)

The notion of mind‐independence plays a central role in the contemporary realism/anti‐realism debate, but the notion is severely ambiguous and consequently the source of considerable misunderstanding. In this paper, four kinds of mind‐independence are distinguished: ontological, causal, structural, and individuative independence. Appreciating these distinctions entails that one can reject the individuative independence of the natural world, and still maintain that the natural world is causally and structurally independent of us. This paper argues that so‐called anti‐realists, especially Rorty, Putnam, and Goodman, (...) are not opposed to the causal and structural independence of the natural world, as is frequently alleged, but rather its individuative independence. An acceptance of these points will hopefully put an end to the prevalence of strawmen in the debate, and focus attention on meatier issues. (shrink)

The capacity to recognise and interpret sluices—bare wh-phrases that exhibit a sentential meaning—is essential to maintaining cohesive interaction between human users and a machine interlocutor in a dialogue system. In this paper we present a machine learning approach to sluice disambiguation in dialogue. Our experiments, based on solid theoretical considerations, show that applying machine learning techniques using a compact set of features that can be automatically identified from PoS markings in a corpus can be an efficient tool to disambiguate (...) between sluice interpretations. (shrink)

Is disability disadvantageous? Although many assume it is paradigmatically so, many disabled individuals disagree. Whom should we trust? On the one hand, pervasive mistrust of already underrepresented groups constitutes a serious epistemic injustice. Yet, on the other, individuals routinely adapt to mistreatment and deprivation and claim to be satisfied. If we take such “adaptive preferences” (APs) at face value, then injustice and oppression may not be recognized or rectified. Thus, we must achieve a balance between taking individuals’ preferences and self-assessment (...) as definitive, and ignoring them entirely. Furthermore, current accounts of APs suffer from an ambiguity: are APs an unreliable guide to individuals’ interests, or to just policy? This chapter argues that we should distinguish between those that are unreliable in the former sense (“well-being APs”), and the latter (“justice APs”). Although all APs are nonautonomous, only well-being APs are irrational. Thus, preferences may be diagnosed as unreliable from the perspective of justice without impugning individuals’ rational capacities. (shrink)

Race theorists have been unable to reach a consensus regarding the basic historical question, “is ‘race’ modern?” I argue that this is partly because the question itself is ambiguous. There is not really one question that race scholars are answering, but at least six. First, is the concept of race modern? Second, is there a modern concept of race that is distinct from earlier race concepts? Third, are “races” themselves modern? Fourth, are racialized groups modern? Fifth, are the means and (...) methods associated with racialization modern? And sixth, are the meanings attached to racialized traits modern? Because these questions have different answers, the debate about the historical origins of “race” cannot be resolved unless they are distinguished. I will explain the ways in which “race” is and is not modern by answering these questions, thereby offering a resolution to a seemingly intractable problem. (shrink)

We present a preliminary study on disambiguation of symbolic expressions in mathematical documents. We propose to use the natural language within which the expressions are embedded to resolve their semantics. The approach is based on establishing a similarity between the expression’s discourse context and a set of terms from Term Clusters based on OpenMath Content Dictionaries. The Term Clusters are semi-automatically constructed terminological resources which classify related mathematical concepts into groups. Each group is labelled with a term which represents (...) the common denominator between the concepts. (shrink)

The conventional history of the ancient Near East at large, including Egypt and the eastern Mediterranean basin, contains several “Dark Ages,” poorly documented transitional periods of uncertain length. James et al. 1991 have argued that the most significant of these Dark Ages—the transition from the Late Bronze to the Iron Age during the last two centuries of the second millennium BCE—is largely an artifact of an overly long reconstruction of the Egyptian Third Intermediate Period, and that this Dark Age presents (...) itself in every chronology linked to the Egyptian. A wide variety of often seemingly contradictory scientific, archaeological, art-historical, and philological evidence has been adduced to argue for the status quo, or to lengthen or to shorten by up to several centuries the relative chronology for this period. This review article comments on the papers presented at the Third BICANE Colloquium held at Sidney Sussex College, Cambridge, summarizing a variety of recent positions taken in the ongoing evaluation of James et al. 1991, testing the specific proposition that the tenth century BCE biblical Egyptian King Shishak is to be distinguished from the historical Egyptian King Shoshenq I, who is now to be situated about a century later in the latter half of the ninth century BCE. the wider historical implications of this proposed distinction are examined in new detail. (shrink)

The problems of access—retrieving linguistic structure from some mental grammar —and disambiguation—choosing among these structures to correctly parse ambiguous linguistic input—are fundamental to language understanding. The literature abounds with psychological results on lexical access, the access of idioms, syntactic rule access, parsing preferences, syntactic disambiguation, and the processing of garden‐path sentences. Unfortunately, it has been difficult to combine models which account for these results to build a general, uniform model of access and disambiguation at the lexical, idiomatic, (...) and syntactic levels. For example, psycholinguistic theories of lexical access and idiom access and parsing theories of syntactic rule access have almost no commonality in methodology or coverage of psycholinguistic data. This article presents a single probabilistic algorithm which models both the access and disambiguation of linguistic knowledge. The algorithm is based on a parallel parser which ranks constructions for access, and interpretations for disambiguation, by their conditional probability. Low‐ranked constructions and interpretations are pruned through beam‐search; this pruning accounts, among other things, for the garden‐path effect. I show that this motivated probabilistic treatment accounts for a wide variety of psycholinguistic results, arguing for a more uniform representation of linguistic knowledge and for the use of probabilistically‐enriched grammars and interpreters as models of human knowledge of and processing of language. (shrink)

We demonstrate “economies of experience” in eye-movement patterns—that is, optimization of eye-movement patterns aimed at more efficient and less costly visual processing, similar to the priming-induced formation of sparser cortical representations or reduced reaction times. Participants looked at Mooney-type, degraded stimuli that were difficult to recognize without prior experience, but easily recognizable after exposure to their undegraded versions. As predicted, eye-movement dispersion, velocity, and the number of fixations decreased with each stimulus presentation. Further analyses showed that this effect was contingent (...) on recognition, and the selection of information from the stimulus could be informed by the identity of the presented object. Finally, our study demonstrates that after exposure to the undegraded version of the stimulus, eye-movement patterns associated with its degraded and undegraded versions become more similar. This suggests that eye-movement patterns can evolve to facilitate the optimal processing of a given stimulus via experience-driven perceptual learning. (shrink)

Adults reason by exclusivity to identify the meanings of novel words. However, it is debated whether, like children, they extend this strategy to disambiguate other referential expressions. To further inform this debate, this study tested 41 adults on four conditions of a disambiguation task: label/label, fact/fact, label/fact, and fact/label. Participants also provided a verbal explanation for their referent selections to tease apart the underlying processes. Results indicated that adults successfully discerned the target object in the label/label and label/fact condition, (...) yet not the remaining two conditions. Verbal reports indicated that the strategy utilized to disambiguate differed depending upon communicative context. These findings confirm that the tendency to reason by exclusivity becomes restricted to word-learning situations with growing linguistic and communicative experience. (shrink)

We present a processing model that integrates same important psychological claims about the human sentence‐parsing mechanism: namely, that processing is influenced by limitations an working memory and by various syntactic preferences. The model uses time‐constraint information to resolve conflicting preferences in a psychologically plausible way. The starting paint far this proposal is the Sausage Machine model (Fodor & Frazier, 1980: Frazier & Fodor, 1978). From there, we attempt to overcome the original model's dependence an ad hoc aspects of its grammar, (...) and its omission of verb‐frame preferences. We also add mechanisms far lexical disambiguation and semantic processing in parallel with syntactic processing. (shrink)

We present a processing model that integrates same important psychological claims about the human sentence‐parsing mechanism: namely, that processing is influenced by limitations an working memory and by various syntactic preferences. The model uses time‐constraint information to resolve conflicting preferences in a psychologically plausible way. The starting paint far this proposal is the Sausage Machine model (Fodor & Frazier, 1980: Frazier & Fodor, 1978). From there, we attempt to overcome the original model's dependence an ad hoc aspects of its grammar, (...) and its omission of verb‐frame preferences. We also add mechanisms far lexical disambiguation and semantic processing in parallel with syntactic processing. (shrink)