Brain and Language Lab

Procedural Deficit Hypothesis

We have proposed that a number of neurodevelopmental disorders can be partially or largely explained by abnormalities of brain structures underlying procedural memory (Ullman, 2004, Ullman and Pierpont, 2005, Evans and Ullman, 2016, Ullman, Earle, Walenski, Janacsek, 2020). According to this neuroanatomical account – the Procedural Deficit Hypothesis, or PDH – these abnormalities should, not surprisingly, lead to the dysfunction of procedural memory. Importantly, since various other functions such as grammar, reading, and math seem to depend on procedural memory, such brain abnormalities may also lead to problems in these domains. We have suggested that the PDH may help account for core symptoms in a variety of neurodevelopmental disorders, including difficulties with language, especially grammar, in specific language impairment and perhaps autism, reading problems in dyslexia, and math impairments in math disability, including dyscalculia.

Since the brain structures underlying procedural memory also play roles in functions other than procedural memory, abnormalities of these structures should also lead to problems with such non-procedural functions. For example, frontal/basal-ganglia circuits subserve not only procedural memory, but a wide variety of other functions, including working memory, attention, and lexical/semantic recall. The frontal/basal-ganglia abnormalities that we have posited and that seem to be found in specific language impairment can thus also explain difficulties with these other functions in the disorder, even though these do not seem to depend on procedural memory.

We have proposed the PDH as an explanatory account for specific language impairment (SLI), dyslexia, motor-speech disorders, autism, Tourette syndrome, ADHD, and mathematical disability. The explanatory power of the PDH is posited to vary across the disorders. For example, it seems to have a large amount of explanatory power in SLI, perhaps less in dyslexia, and less in autism. Moreover, the nature of the abnormalities and thus the ensuing procedural memory and other dysfunctions are posited to differ across the disorders. Indeed, in some cases, the abnormalities may manifest as superior procedural memory abilities, as compared to typically developing children. This may be the case, for example, in Tourette syndrome (Walenski et al., 2007; Dye et al., 2016; Takács et al., 2017; Takács et al., 2018) and perhaps in (some populations of) autism (Walenski et al., 2014).

Importantly, which brain structures underlying procedural memory are affected, the nature of their abnormalities, and what additional structures are involved, are all posited to vary across the disorders, thus helping explain the differences between them. Nevertheless, we suggest that comorbidities among certain of these disorders can be partly if not largely explained by their common brain abnormalities (Ullman and Pierpont, 2005, Evans and Ullman, 2016).


We have also proposed that declarative memory compensates for procedural memory and other dysfunctions in a wide range of disorders, including SLI, dyslexia, autism, Tourette Syndrome, and OCD (Ullman and Pullman, 2015a; Ullman and Pullman, 2015b). According to this view (the Declarative Memory Compensation Hypothesis), declarative memory remains largely normal in these and various other disorders, and, because it is very flexible, it can and often does learn a wide range of strategies to at least partially overcome numerous types of impairments. In some cases declarative memory may even be enhanced as compared to typically developing children, including in dyslexia (Hedenius et al., 2013), specific language impairment (Lukacs et al., 2017), and autism (Walenski et al., 2008), thus in turn enhancing its ability to compensate.

For example (See Ullman and Pullman, 2015), declarative memory seems to help people with OCD or Tourette syndrome learn to control compulsions and tics; it allows individuals with autism to compensate for social deficits by memorizing scripts for navigating social situations; and it learns strategies to overcome reading and language difficulties in people with SLI (Lum et al., 2012, Conti-Ramsen et al., 2015), dyslexia (e.g., Hedenius et al., 2013), and autism.

Declarative memory compensation has potentially important clinical implications. First, designing treatments that rely on declarative memory, or that improve learning in this system, could enhance compensation and thus reduce symptoms. For example, drugs known to improve declarative memory could ameliorate symptoms, particularly if paired with behavioral therapies learned in this system.

Second, developing treatments that avoid compensation by declarative memory may increase the likelihood of stimulating and potentially strengthening the dysfunctional circuitry. This is akin to therapies that prevent patients from using their good hand so they are forced to use their bad one, which can improve its functionality.

The compensation also has implications for diagnosis. Many individuals might be compensating their way out of clinical diagnosis (the Compensation Underdiagnosis Hypothesis; Ullman and Pullman, 2015). This should hold especially for individuals or groups with better declarative memory abilities. Indeed, underdiagnosis due to compensation may help explain one of the major conundrums of all these disorders: why they are diagnosed more commonly in boys than girls. On this view, declarative memory seems to be better, on average, in girls and women than boys and men. Thus females are likely to compensate more successfully than males, even compensating themselves out of diagnosis more often than males (Ullman and Pullman, 2015).

Compensation by declarative memory may also be found in other disorders (Ullman and Pullman, 2015), including ADHD and debilitating adult-onset disorders such as aphasia (Drury and Ullman, 2002) and Parkinson’s disease (Ullman et al., 2002). Indeed, given the power and flexibility of declarative memory, it may play compensatory roles in a wide range of disorders.