Dopaminergic mechanisms of inference for language: a pharmaco-fMRI study

Humans are flexible and efficient at inferring meaning during language
processing. Similar capacity has also been observed in making novel decisions
drawing from past experience, for example in spatial relations.

Dopamine is the key neural candidate posited to play a crucial role in drawing
inferences in non-linguistic domains. Dopamine is a catecholamine
neurotransmitter that is known to play a central role in flexible and
self-directed thought and action: our abilities to think about and make plans
based on stimuli that are not physically present (working memory), to learn
from new information and stimulus-response associations (reinforcement
learning), and to make choices based on prior and current environment
(incentive motivation), all critically rely on dopamine. Particularly, dopamine
stimulation leads to flexible and adaptive behaviour, making it a likely
component for inferential reasoning.

Besides its role in flexible behaviour, dopamine has also been linked to
language processing: previous research has shown the importance of dopamine in
amplifying the salience of linguistic information (e.g. enhancing semantic
processing). This is consistent with the idea that dopamine increases
signal-to-noise ratio between specific signals and background noise.

This study addresses fundamental questions about the neural mechanisms that are
central to our capacity to infer meanings flexibly and efficiently during
language processing, broadening the role of dopamine from action and spatial
planning to linguistic relations.

The primary objective of this study is to investigate whether increases in
brain dopamine enhance novel meaning inference in language processing, and what
cognitive and neural mechanisms underlie this. We will test the hypothesis that
dopamine promotes the building of novel word meaning representations and
increases the signal to noise ratio of existing word meaning representations in
the hippocampus and prefrontal cortex.

A double-blind placebo-controlled between-subject design will be employed:
Healthy participants are tested once, either on placebo or on a low oral dose
(150mg) of the dopaminergic precursor, levodopa, in order to increase brain
dopamine. Similar pharmacological designs, albeit with other dopaminergic
pharmacological challenges, are commonly used in our lab. The proposed dose of
levodopa has been widely used without side effects in other studies with
comparable study populations.

Participants will attend two study sessions: a screening session and a
pharmaco-fMRI session (levodopa or placebo). On the pharmaco-fMRI session,
participants will complete questionnaires, structural and functional MRI scans,
as well as a battery of tasks both in (during the fMRI scans) and outside the
scanner. On the day preceding each pharmaco-fMRI session, participants will
have to adhere to some simple restrictions with respect to medication, alcohol
and drug intake. On the pharmaco-fMRI day, subjects will have to refrain from
smoking and stimulant-containing drinks. Levodopa can be administered safely
without any relevant risk of serious adverse events and has been approved for
clinical use in the Netherlands.