Alcohol Drinking

Abstract

Alcohol drinking is a type of drinking behavior: the consumption of alcoholic beverages, which cognitive psychology studies less as a pharmacological event than as a behavior governed by what the drinker expects, perceives, and automatically does. This article sets out alcohol expectancy theory, the balanced-placebo design that separates what a drink does from what it is believed to do, the acute narrowing of attention and memory that intoxication produces, and the implicit, dual-process cognition that drives heavy use. Expectancies learned before the first drink shape its social effects; the pharmacological effect disrupts the appraisal of stress and encodes memory poorly; and automatic approach tendencies, once learned, can be retrained. Three demonstrations make alcohol expectancies, the balanced-placebo logic, and dual-process control concrete.

Keywords: alcohol drinking, balanced-placebo design, alcohol expectancies, implicit cognition

Alcohol drinking is the heading the National Library of Medicine uses for the consumption of alcoholic beverages, and in cognitive psychology it is a case in which belief and pharmacology are unusually hard to tell apart. A drink changes the body, but it also arrives wrapped in a lifetime of learned expectation about what it will do, and much of what looks like a drug effect turns out, on inspection, to be an effect of the expectation. The cognitive science of drinking is therefore largely an effort to pull those two causes apart: to ask what alcohol does to attention, appraisal, and memory, and what the belief that one has drunk does on its own. The answers reach from the social expectancies formed before a first glass to the automatic, implicit processes that keep heavy drinking going against intention.

What Alcohol Drinking Is

Within the Medical Subject Headings, Alcohol Drinking sits beneath drinking behavior as the consumption of alcoholic beverages specifically, the branch of drinking in which the fluid is alcoholic and therefore psychoactive. The relation to the parent is one of specification: drinking behavior is the general act of ingesting liquids and the attitudes surrounding it, and alcohol drinking is the member of that category whose cognitive interest comes from the drug it delivers. What distinguishes alcohol drinking from drinking in general is not the motor act but the fact that the beverage alters the very cognitive systems, attention, appraisal, and memory, that would otherwise regulate it.

That circularity is the core of the subject. A drink is chosen by a mind and then acts upon that mind, so the behavior and its pharmacology are entangled in a way that ordinary ingestion is not. The reason alcohol drinking has a rich cognitive psychology, rather than only a pharmacology, is that a large share of its observable effects flow not from the ethanol but from the drinker’s learned expectations about it, and separating the two has been the organizing methodological problem of the field. The sections that follow trace that separation: first the expectancies that precede the drink, then the design that isolates them, then the drug’s genuine effects on cognition, and finally the automatic processes that sustain heavy use.

Types of Alcohol Drinking

Medical Subject Headings places three narrower descriptors beneath its parent heading drinking behavior under Alcohol Drinking, each naming a pattern of consumption distinguished by its quantity, timing, or population rather than by a different underlying psychology. The subtypes are a classification of drinking patterns, not a partition of cognitive mechanisms: the expectancy, attentional, and dual-process processes described in this article operate across all of them, and a single drinker may fall under more than one. MeSH is an indexing vocabulary, so the categories reflect how the literature is catalogued rather than a claim that these are the only, or the most theoretically fundamental, kinds of drinking.

Table 1

MeSH subtypes of alcohol drinking and their cognitive focus.

Note. Binge drinking and underage drinking are live topics on this site and are linked; alcohol drinking in college is a MeSH descriptor without a dedicated page here. The subtypes are indexing categories and are not mutually exclusive.

Alcohol Expectancies

The most durable contribution of cognitive psychology to the study of drinking is alcohol expectancy theory: the claim that a person’s beliefs about what alcohol will do, stored in memory and activated when drinking is in prospect, are a primary determinant of whether and how much they drink. The foundational demonstration showed that expectancies of reinforcement from alcohol, that it makes social situations easier, enhances pleasure, increases power, or relaxes tension, form a measurable domain and relate systematically to drinking patterns, with heavier drinkers holding stronger positive expectancies (Brown et al., 1980). Crucially, these beliefs are acquired through social learning before a person has ever drunk, from observation, advertising, and culture, so they are causes of drinking rather than merely its consequences.

The theory was elaborated into a memory-network account on which expectancies are stored as associative structures that a drinking context primes, so that the expected effect is activated automatically and shapes behavior before any pharmacological effect can occur. A systematic review of the expectancy literature concluded that expectancies predict the onset, level, and problems of drinking across the lifespan, and that they mediate much of the influence of more distal risk factors, making them a leverage point for prevention (Jones, Corbin, & Fromme, 2001). The demonstration lets expectancy strength and actual dose vary independently, showing how the expected effect can move behavior on its own.

Expectancy versus dose in a drink’s effect

A drinker’s response combines what they expect alcohol to do with what the dose actually does. For a socially mediated behavior, belief carries most of the weight; for a motor or higher-cognitive task, the pharmacological dose dominates. Vary expectancy strength and dose, and switch the behavior, to see the composition change.

The Balanced-Placebo Design

If expectancies and pharmacology both influence behavior, an experiment must separate them, and the instrument that does so is the balanced-placebo design. It crosses two factors independently: what the participant is told they are drinking (alcohol or not) and what they actually receive (alcohol or not), yielding four cells. Because beverage can be disguised, a vodka-and-tonic mixed and presented so that taste and smell do not betray the content, a person can be given alcohol while believing it is tonic, or tonic while believing it is alcohol. The design grew out of an experimental analogue of loss-of-control drinking in which alcoholics and social drinkers, told they were taste-rating beverages, drank according to what they were told the drinks contained rather than what they actually contained (Marlatt, Demming, & Reid, 1973).

The logic of the four cells is what makes the design powerful: a pharmacological effect appears as a difference between what was received, holding belief constant, and an expectancy effect appears as a difference between what was told, holding dose constant (Rohsenow & Marlatt, 1981). The repeated finding is a dissociation. For many socially mediated behaviors, sociability, aggression, and sexual arousal to stimuli, behavior tracks what the person was told, so that the belief one has drunk produces the effect whether or not one has. For motor coordination and higher cognitive performance, by contrast, behavior tracks the actual dose. The design thus shows that alcohol’s reputation does part of its work, while its pharmacology does the rest.

Decomposing the balanced-placebo design

Set the mean response in each of the four cells and watch the design pull apart the two causes. The expectancy effect is the told-alcohol rows minus the told-tonic rows; the pharmacological effect is the given-alcohol cells minus the given-tonic cells. The defaults reproduce the worked example in the text.

Acute Effects on Cognition

Set expectancy aside and alcohol still has genuine pharmacological effects on cognition, and they are specific rather than a general dulling. The most influential account is alcohol myopia: intoxication narrows the span of attention so that behavior is controlled by the most immediate, salient cues and the inhibiting, more abstract considerations that a sober mind would weigh fall out of view (Steele & Josephs, 1990). Myopia explains why alcohol can both inflame and calm, embolden and excuse: it does not add a fixed effect but amplifies whatever cue is most present, so the same dose makes an angry person angrier and a distracted person carefree.

The same narrowing underlies alcohol’s much-claimed tension reduction. The appraisal-disruption model holds that alcohol relieves stress not by sedating directly but by impairing the cognitive appraisal of threatening information, so that a worry must be actively processed to be felt and alcohol blocks that processing, which is why a distracting activity can substitute for the drink (Sayette, 1993). At higher doses the drug’s effect on memory becomes stark. Alcohol impairs the hippocampal consolidation of new experience into long-term memory while leaving already-stored memory and immediate attention relatively intact, producing the blackout: a period of normal-seeming behavior that leaves no retrievable trace, which in its fragmentary and complete (en bloc) forms is a dissociation of encoding from performance rather than a loss of consciousness (Wetherill & Fromme, 2016).

Implicit Cognition and Dual-Process Control

Expectancy theory describes explicit, reportable beliefs, but much of the cognition that drives heavy drinking is implicit, operating automatically and outside deliberate control. A dual-process account of the alcohol-behavior link holds that drinking is the joint product of a fast, associative system that responds to the learned incentive value of alcohol cues and a slower, reflective system that acts on considered goals, with problem drinking arising when the automatic system is strong and reflective control is weak (Moss & Albery, 2009). This is the same architecture that organizes self-control more generally, a reflective system that can be depleted or outmatched by an impulsive one (Hofmann, Friese, & Strack, 2009).

The automatic system leaves measurable fingerprints. Chief among them is an attentional bias: alcohol-related cues capture and hold the attention of heavy drinkers, a bias that tracks craving and predicts relapse, and that develops through the same associative learning that gives cues their incentive value (Field & Cox, 2008). Because the bias is a learned association rather than a fixed trait, it is a treatment target. Teaching dependent patients to repeatedly push alcohol cues away using a joystick task reverses their automatic approach tendency and measurably improves abstinence a year later, direct evidence that the implicit component is both causal and modifiable (Wiers et al., 2011). The demonstration shows how the balance between an impulsive pull and reflective control determines whether a drink is taken.

The impulsive pull against reflective control

Dual-process accounts cast a drink as the outcome of a contest between a fast, automatic pull toward alcohol cues and a slower, reflective control system. A drink is taken when the pull outweighs control. Raise the impulsive pull, adjust control, and deplete the reflective system to see the balance tip.

Worked Example

Consider how the balanced-placebo design separates the two causes of a social behavior, using illustrative cell means for a measure of sociability (say, the number of sociable remarks in a brief conversation task). Suppose the four cells yield: told tonic and given tonic (control) = 10; told tonic and given alcohol (antiplacebo) = 12; told alcohol and given tonic (placebo) = 18; told alcohol and given alcohol (full) = 20. The expectancy main effect is the difference between the told-alcohol rows and the told-tonic rows, averaging over what was actually received: mean(18, 20) − mean(10, 12) = 19 − 11 = 8. The pharmacological main effect is the difference between the given-alcohol cells and the given-tonic cells, averaging over instruction: mean(12, 20) − mean(10, 18) = 16 − 14 = 2. The expectancy effect of 8 is four times the pharmacological effect of 2, so for this socially mediated behavior belief does most of the work. The interaction, (20 − 18) − (12 − 10) = 2 − 2 = 0, is zero, meaning the two causes simply add. The lesson is that an ordinary drug study comparing only the control and full cells would report a total effect of 20 − 10 = 10 and attribute all of it to alcohol, whereas the full design reveals that four-fifths of that difference is expectancy.

Discussion

Alcohol drinking is a clarifying case for cognitive psychology because it forces a distinction that most behaviors let us ignore: between what a substance does and what the belief about the substance does. The balanced-placebo design made that distinction experimentally tractable, and the result, that socially mediated effects ride largely on expectancy while motor and higher-cognitive impairment ride on dose, reorganized the field around cognition rather than pharmacology alone. Expectancies formed before the first drink, a drug that narrows attention and blocks appraisal and encoding, and an implicit system of biases and approach tendencies together make drinking a problem of mind as much as of chemistry.

Table 2

Cognitive mechanisms in alcohol drinking and what each explains.

Note. Each mechanism is drawn from a section above; together they show drinking as a joint product of expectancy, pharmacology, and automatic cognition rather than of any one alone.

The cognitive account has limits worth stating. It describes the early and social course of drinking well, but the transition to dependence recruits neuroadaptations that shift drinking from positive toward negative reinforcement and organize it around a cycle of binge, withdrawal, and preoccupation that the vocabulary of expectancy and attention captures only partially (Koob & Volkow, 2016).

Current Directions

Two developments are shaping the contemporary field. The first is the maturation of craving as a measurable cognitive construct. A synthesis of the craving literature argues that craving is best understood as a state in which automatic appetitive processes commandeer attention and working memory, and that progress depends on measuring it in real time and in the cue reactivity provoked by alcohol-associated stimuli rather than by retrospective self-report, aligning the construct with the implicit-cognition work above (Sayette, 2016). The second is the integration of cognitive accounts with a neurocircuitry framework in which the implicit biases and expectancies have identifiable substrates in the brain’s reward, habit, and executive systems, so that the stage of the addiction cycle predicts which cognitive process dominates (Koob & Volkow, 2016).

Running through both is the translation of mechanism into intervention. Having shown the automatic approach tendency to be causal and reversible (Wiers et al., 2011), researchers are refining cognitive bias modification and combining it with expectancy-challenge and appraisal-based approaches, so that prevention and treatment increasingly act on identified cognitive processes rather than on drinking in the abstract.

Cognitive Implications

Read as cognition, alcohol drinking illustrates three themes that reach beyond addiction. First, it is the clearest everyday demonstration that belief is causal: the placebo and antiplacebo cells of the balanced-placebo design show an expectation producing a behavioral effect in the absence of its supposed physical cause, a result that bears on the general psychology of suggestion and the top-down shaping of experience (Brown et al., 1980; Rohsenow & Marlatt, 1981). Second, alcohol myopia offers a natural manipulation of attentional scope, showing how much of what we call judgment depends on the breadth of what is attended, and how behavior changes when that breadth is pharmacologically narrowed (Steele & Josephs, 1990). Third, the dissociation of encoding from performance in the blackout is a striking case of the independence of memory systems, a period of competent, attentive behavior that leaves no episodic trace (Wetherill & Fromme, 2016). In each, drinking serves as an unusually legible window onto a general cognitive process.

Common Misconceptions

Alcohol’s social effects are purely pharmacological.

The balanced-placebo design shows that for many socially mediated behaviors, sociability, aggression, and sexual arousal among them, the behavior tracks what the person believes they drank rather than what they actually drank. Much of alcohol’s reputation is an expectancy effect (Rohsenow & Marlatt, 1981).

A blackout means passing out.

A blackout is a failure to form new long-term memories, not a loss of consciousness. A person in a blackout can walk, talk, and act normally; the experience simply leaves no retrievable trace, because alcohol has blocked its consolidation (Wetherill & Fromme, 2016).

Heavy drinking is simply a failure of willpower.

Much of what sustains heavy drinking is implicit: an automatic attentional bias and approach tendency toward alcohol cues that operate beneath deliberate choice. Because these are learned associations, they can be retrained directly, which willpower alone does not address (Field & Cox, 2008; Wiers et al., 2011).

Glossary

Alcohol drinking.

The consumption of alcoholic beverages; in cognitive psychology, a behavior whose effects are jointly determined by the pharmacology of ethanol and by the drinker’s learned expectations about it.

Alcohol expectancy theory.

The account on which beliefs about alcohol’s effects, learned socially before drinking and stored in memory, are a primary cause of whether and how much a person drinks.

Alcohol myopia.

The acute narrowing of attention under intoxication so that behavior is controlled by the most immediate, salient cues while more abstract, inhibiting considerations are not processed.

Antiplacebo condition.

The balanced-placebo cell in which a person is given alcohol but told they are not, isolating the pharmacological effect from the expectancy.

Appraisal-disruption model.

Sayette’s account on which alcohol reduces stress by impairing the cognitive processing of threatening information rather than by sedating directly, so that distraction can substitute for the drink.

Approach bias.

An automatic action tendency to move toward alcohol-related stimuli; measurable, predictive of heavier drinking, and reversible by cognitive bias modification.

Attentional bias.

The preferential capture and holding of attention by alcohol-related cues; it tracks craving and predicts relapse.

Balanced-placebo design.

An experimental design crossing what a participant is told they are drinking with what they actually receive, allowing pharmacological and expectancy effects to be separated.

Binge drinking.

A pattern of heavy episodic drinking that raises blood alcohol concentration to intoxicating levels in a short span; a MeSH subtype of alcohol drinking.

Blackout.

An alcohol-induced failure to consolidate new experience into long-term memory, fragmentary or complete (en bloc), occurring despite apparently normal behavior at the time and without loss of consciousness.

Cue reactivity.

The set of physiological, subjective, and attentional responses provoked by alcohol-associated cues, through which learned associations influence drinking.

Dual-process model.

The account on which drinking is the joint product of a fast, automatic appetitive system and a slower, reflective control system, with problem drinking arising when the first is strong and the second weak.

Implicit cognition.

The automatic, largely non-conscious associative processes, such as attentional bias and approach tendencies, that influence drinking independently of explicit, reportable beliefs.

Loss-of-control drinking.

The claim that one drink triggers uncontrollable further drinking in dependent people; the balanced-placebo analogue showed the effect tracks belief about the beverage rather than its actual content.

Key Researchers

Matt Field

. Professor of Psychology at the University of Sheffield; his work on attentional bias, implicit cognition, and dual-process accounts has helped establish that automatic processes drive much of addictive behavior and can be measured and retrained. ORCID · Google Scholar

Kim Fromme

. Professor Emeritus of Psychology at the University of Texas at Austin; she studies alcohol expectancies, behavioral risk-taking, and alcohol-induced blackouts, and co-authored a major review of expectancy theory. Google Scholar

Mark S. Goldman

. Distinguished Professor Emeritus of Psychology at the University of South Florida; a pioneer of alcohol expectancy theory and the memory-network model, he established expectancies as measurable antecedents of drinking. Faculty page

G. Alan Marlatt

. Addiction researcher at the University of Washington (1941-2011); he designed the balanced-placebo paradigm and the loss-of-control-drinking analogue that separated the belief in alcohol from its pharmacology. Wikipedia

Michael A. Sayette

. Distinguished Professor of Psychology at the University of Pittsburgh; his appraisal-disruption model and his work on craving and cue reactivity clarified how alcohol acts on the processing of stress and how craving loads attention. Google Scholar

Frequently Asked Questions

What is alcohol drinking in cognitive psychology?

It is the consumption of alcoholic beverages, the sense in which the National Library of Medicine uses the term, studied not mainly as a pharmacological event but as a behavior shaped by what the drinker expects, perceives, and automatically does. Its distinctive feature is that the drink alters the very cognitive systems, attention, appraisal, and memory, that would otherwise regulate it (Brown et al., 1980).

What is alcohol expectancy theory?

It is the account on which beliefs about what alcohol will do, learned socially before a person has ever drunk and stored in memory, are a primary cause of drinking. These expectancies predict the onset, level, and problems of drinking, and because they precede experience they are causes rather than consequences (Brown et al., 1980; Jones, Corbin, & Fromme, 2001).

What is the balanced-placebo design?

It is an experiment that crosses what a person is told they are drinking with what they actually receive, producing four conditions. Because beverage content can be disguised, the design separates the pharmacological effect of alcohol from the effect of merely believing one has drunk (Rohsenow & Marlatt, 1981).

Does alcohol cause its social effects pharmacologically?

Often not. For many socially mediated behaviors, sociability, aggression, and sexual arousal, the balanced-placebo design shows behavior tracks what the person was told they drank rather than the actual content. Motor and higher-cognitive impairment, by contrast, track the real dose (Marlatt, Demming, & Reid, 1973).

What is alcohol myopia?

It is the acute narrowing of attention under intoxication so that behavior is controlled by the most immediate, salient cues while more abstract, inhibiting considerations drop out of view. Myopia explains why alcohol amplifies whatever cue is most present rather than adding a single fixed effect (Steele & Josephs, 1990).

Why do people experience blackouts?

Because alcohol, at higher doses, impairs the hippocampal consolidation of new experience into long-term memory while leaving immediate attention and stored memory relatively intact. The result is a period of normal-seeming behavior that leaves no retrievable trace, in fragmentary or complete form (Wetherill & Fromme, 2016).

What is attentional bias for alcohol?

It is the automatic capture and holding of a heavy drinker’s attention by alcohol-related cues, a bias that tracks craving and predicts relapse. It develops through associative learning that gives cues their incentive value, and it is a measurable, modifiable marker of implicit cognition (Field & Cox, 2008).

Can automatic drinking tendencies be changed?

Yes. Training dependent patients to repeatedly push alcohol cues away reverses their automatic approach tendency and improves abstinence a year later, showing the implicit component is causal and can be retrained directly rather than merely resisted (Wiers et al., 2011).

References