General and Theoretical Psychology

Abstract. General and theoretical psychology investigates the conceptual architecture of psychological knowledge: what counts as a psychological phenomenon, how psychological constructs are defined, which kinds of evidence support claims about them, and how explanations can connect experience, behavior, development, social relations, computation, and biological organization. The field integrates the findings of specialized psychological sciences while examining the assumptions that make their theories intelligible and testable. This monograph develops a systematic account of psychological explanation from the historical emergence of experimental psychology through contemporary questions about construct validity, causal inference, multilevel theory, reproducibility, and generalization across populations and settings.

The analysis draws on original historical studies, influential theoretical investigations, experimental programs, psychometric research, large-scale replications, and methodological syntheses. Its independent contribution is a proposed architecture of conditional psychological explanation: explanations are made explicit through six linked commitments concerning phenomenon, construct, mechanism, context, evidence, and scope. This architecture makes competing theories comparable, exposes untested extrapolations, and supplies a research protocol for assessing psychological claims across biological and potentially artificial cognitive systems. The proposal is a theoretical synthesis advanced in this monograph, not a separately validated empirical discovery.

The Scientific Subject of General and Theoretical Psychology

General psychology concerns the principles, processes, and regularities through which psychological phenomena are described and investigated across specialized domains. Theoretical psychology evaluates the concepts, explanatory structures, methodological commitments, and ontological assumptions of psychological science. Together they form an intellectual center for integrating results that otherwise remain dispersed among research traditions. They ask how perceptual organization relates to learning, how action relates to representation, how individual differences interact with experimental effects, why psychological measurements acquire meaning, and under which conditions a laboratory finding speaks about life beyond the laboratory.

The basic unit of inquiry is a psychological claim: a proposition concerning an identifiable process, state, disposition, relation, or pattern of behavior, accompanied by assumptions about how it is observed or inferred. The concept of attention may refer to selective processing, task engagement, a computational allocation policy, a family of measurable performance differences, or a particular neural network's activity. A theoretically adequate definition specifies which of these is at issue and establishes the relations among them. The same word can identify multiple constructs; different words can refer to overlapping operations. Conceptual precision is therefore an empirical resource, because it determines which comparisons are possible.

Psychological phenomena are simultaneously temporal and relational. Remembering requires an organism or system with a history, an opportunity for encoding and retrieval, and a criterion of successful retention. Motivation concerns the organization and pursuit of goals in a context where alternative actions are possible. Emotion involves patterns of appraisal, experience, physiology, and action whose relations vary with theoretical definition and measurement strategy. Social identity concerns processes whose enabling conditions include cultural categories and interactional structures. General psychology investigates these relations without presupposing that a single descriptive vocabulary is sufficient for every phenomenon.

Phenomenon, construct, measure, and model

A phenomenon is a pattern or event that inquiry seeks to understand. A construct is a theoretical designation used to organize observations and explanatory propositions. An operation is the specified procedure through which a variable is manipulated, measured, or classified. A model is an articulated representation of relations among variables, mechanisms, or processes. An explanation relates the target phenomenon to enabling conditions, causal structures, functional requirements, or principled dependencies. These entities interact, but their identities remain analytically distinct.

Consider working memory. A delayed-response task yields behavior; accuracy and response latency become operational measures; a model specifies storage, maintenance, transformation, or attentional control; theoretical explanations propose mechanisms that produce the observed task-dependent performance. Agreement with one task cannot alone establish that a model describes every component of working memory. The general-theoretical task is to assess convergence across operations, boundary conditions, rival mechanisms, and individual or cultural contexts.

Description, prediction, explanation, and intervention

Description organizes observations into stable discriminations. Prediction estimates outcomes from known information. Explanation specifies why patterns occur under stated conditions. Intervention alters candidate causes and tests their consequences. Predictive success is useful evidence for some propositions, but a flexible algorithm can predict an outcome from correlates without identifying the process that produces it. A causal intervention can establish a local relationship while leaving broader mechanisms unsettled. General psychology assesses what a design licenses and which further inferences require new evidence.

This distinction has practical consequences. A questionnaire may predict later behavior while its underlying construct remains uncertain. A neuroscience classifier may separate experimental conditions while providing limited evidence about the psychological process encoded. A computational model may reproduce a behavioral curve through several alternative algorithms. Each result has scientific value when its inferential reach is declared and independently tested.

The Historical Formation of Psychological Explanation

Psychophysics and experimental institutions

The modern disciplinary form of psychology emerged through several converging inquiries into sensation, physiology, timing, learning, and the relation between conscious experience and measurable events. Ernst Weber's investigations of sensory discrimination and Gustav Fechner's psychophysics demonstrated that carefully structured comparisons could connect variations in physical stimulation to patterns of reported perception. Such research established a durable methodological question: how can experiences that are publicly inaccessible in the first person yield reliable relations to publicly controlled stimuli and responses?

In 1879 Wilhelm Wundt established the psychology institute at Leipzig University, an institution widely treated as a foundational moment of organized experimental psychology. The laboratory brought controlled observation, instrumentation, training, and methodological debate into an academic program. Its significance lies in the consolidation of research practices rather than the sudden invention of all psychological questions. Psychological inquiry also drew on earlier philosophy, physiology, medicine, education, anthropology, and traditions outside the European academy. The institutional history of modern psychology should be distinguished from the broader history of human theories of mind. Leipzig University's institutional history documents the experimental foundation and antecedents.

Hermann Ebbinghaus extended quantitative experimental work to memory. In Über das Gedächtnis (1885), he examined acquisition, retention, relearning, and forgetting through repeated study of verbal material. His initial extensive experimentation on himself made it possible to investigate regularities within a person with carefully controlled material, while also presenting a vivid limit on population generalization. The original author's preface expressly identifies this limitation. His contribution extends beyond the famous forgetting curve: memory became a dynamic process that could be experimentally decomposed, measured across delays, and explained through systematic manipulations.

Experience and functional organization

William James's The Principles of Psychology (1890) brought together habit, attention, will, emotion, self, and the stream of consciousness in an encompassing treatment of mental life. Its lasting theoretical importance includes the view that mental processes unfold in time and that psychological categories acquire meaning through the ways organisms adapt and act. Functional questions ask what a process accomplishes and under which circumstances its activity makes a difference. They supplement explanations of internal organization with questions about the environment and the purposes of action.

The early twentieth century expanded psychology through competing programs. Gestalt investigators examined perceptual organization and argued that relational structure is essential to explaining what observers perceive. Kurt Koffka's Principles of Gestalt Psychology (1935) illustrates the breadth of this program, linking perception, learning, memory, and the environmental field. Behaviorist traditions demanded disciplined attention to observable response patterns and the conditions that alter them. Later theoretical disputes revealed that an exclusively stimulus–response vocabulary could be insufficient to characterize learned relations, planning, and internal representation.

Learning, representation, and social mediation

Edward Tolman's “Cognitive Maps in Rats and Men” (1948) provided an influential formulation of how spatial and goal-directed learning may be described in representational terms. The concept of a cognitive map offered an explanatory structure richer than a fixed chain of reinforced responses. The particular empirical results and later neuroscientific extensions require independent evaluation, but Tolman's larger methodological contribution remains clear: observable behavior can motivate hypotheses about internally organized relations that themselves require testable predictions.

Frederic Bartlett's Remembering (1932) treated memory as reconstructive and situated in existing knowledge and social meaning. The account broadened memory theory beyond isolated retention scores by considering how remembered material is transformed in culturally meaningful tasks. Its historical influence does not imply that every reconstruction claim is confirmed by modern experiments. Bartlett supplies a model of how naturalistic observation and interpretive description can generate hypotheses to be evaluated through varied measures.

Developmental and cultural-historical approaches situated psychological functioning within changing forms of action and social participation. Jean Piaget's investigations developed influential models of the transformation of cognitive structures across childhood. Lev Vygotsky's cultural-historical framework emphasized the mediated development of higher psychological functions through language, tools, and social relations. These programs introduced explanatory time scales and relational conditions frequently lost in brief laboratory protocols. Theoretical integration requires placing their claims alongside contemporary longitudinal, comparative, and intervention evidence, rather than treating entire schools as indivisible propositions.

Albert Bandura, Dorothea Ross, and Sheila Ross's 1961 experiment on imitation of aggressive models demonstrated that observational learning could be investigated experimentally through defined models and measured child behavior. It did not establish a universal law connecting every observed aggressive scene to later aggression. Its enduring significance is that psychological explanation must sometimes include another agent's behavior as a condition of learning. Social models, expectations, opportunities, and situational cues become part of causal research, extending theoretical psychology beyond isolated individual response.

The cognitive turn and contemporary plurality

The postwar development of information-processing approaches made it increasingly common to describe cognition in terms of representations, transformations, goals, decisions, and memory operations. Experimental tasks offered constrained behavioral evidence for such accounts; computer simulation supplied a way to turn verbal hypotheses into procedures with observable consequences. This development transformed, but never eliminated, disputes about embodiment, environment, intentionality, consciousness, and the explanatory role of social organization.

Psychology now contains biological, cognitive, behavioral, phenomenological, developmental, cultural, evolutionary, social, and computational research programs. Their plurality is a productive fact when theories articulate compatible or competing causal claims and submit them to appropriate tests. Problems arise when a vocabulary developed for one purpose is assumed to govern every other purpose. General and theoretical psychology provides the comparative framework required to state when the programs address the same phenomenon, when they explain different aspects, and when their predictions genuinely conflict.

Major Traditions in Psychological Theory

Behavioral and learning explanations

Behavioral theories analyze systematic relations among environments, histories of reinforcement, discriminative conditions, and action. Their explanatory strength lies in the precise organization of observable contingencies and in experimental procedures that permit the direct manipulation of environmental conditions. Such procedures have produced robust principles of associative learning across species and settings. Nevertheless, translating a laboratory contingency into a general theory of human action requires attention to language, prior learning, competing incentives, social constraints, and the organism's own history.

The theoretical distinction between a procedure and a mechanism is essential. Classical conditioning describes an experimental arrangement and acquired relations among stimuli and responses. It does not, by itself, identify every neural or cognitive process that produces learning. An explanatory model may use associative strength, uncertainty reduction, temporal prediction, or representational updating. Competing models must be tested through conditions under which they predict different learning trajectories, extinction patterns, or transfers to new contexts.

Cognitive and representational explanations

Cognitive theories identify internal structures and operations from patterned performance. They explain why a particular manipulation changes recall, attention, classification, or judgment by positing a functional organization that generates the difference. Representation is therefore a hypothesis about causal organization, not simply a synonym for unseen mental content. Strong cognitive models derive discriminating predictions concerning errors, timing, transfer, interference, and interactions among task requirements.

Amos Tversky and Daniel Kahneman's 1974 analysis of judgment under uncertainty examined representativeness, availability, and anchoring as principles relevant to observed judgments. Their work exposed regular patterns that could be investigated through carefully constructed decision problems. Subsequent criticism and alternative ecological accounts advanced the field by asking how a proposed heuristic is specified, what information its user possesses, and which environmental distributions make a procedure accurate or misleading. The resulting research agenda concerns mechanisms and boundary conditions rather than a permanent division between rational and irrational people.

Biological, dynamical, and ecological explanations

Biological psychology connects behavioral and experiential patterns to neural, endocrine, developmental, and bodily processes. Such connections depend on the explanatory level: a neural correlate of task performance is evidence of association, while a well-designed lesion, stimulation, pharmacological manipulation, or natural experiment may support particular causal claims. The existence of a neural substrate is a background condition for human psychological functioning; a successful biological explanation still has to specify how its proposed variables generate the phenomenon.

Dynamical-systems accounts model change through coupled variables evolving across time. Ecological approaches emphasize the structure of organism–environment relations and the information available for action. These frameworks often challenge assumptions that a static representation inside an individual is the unique explanatory unit. Their contribution is strongest when their state variables, environmental constraints, and time-dependent predictions are explicitly operationalized. A formal dynamical account is subject to model comparison like any other theoretical proposal.

Developmental, social, and cultural explanations

Psychological mechanisms develop through biological maturation, learning, institutions, interaction, and historical participation. A developmental model asks how capabilities reorganize across repeated experiences rather than presuming that an adult experimental effect is identical in childhood. Social and cultural theories investigate the conditions through which categories, norms, language, relationships, and material arrangements shape psychological organization. These approaches improve theory when context is specified as a variable or causal condition, rather than invoked only after an unexpected result.

Joseph Henrich, Steven Heine, and Ara Norenzayan's 2010 critique of WEIRD sampling established a major methodological challenge to claims about universal human psychology based on narrow participant populations. Its central implication for theory is conditional scope: findings obtained in one cultural and institutional setting warrant further tests before being generalized to the species. Cross-cultural research similarly requires credible measurement equivalence; translated questionnaire items alone do not establish that the same latent construct has the same meaning or psychometric organization in every population.

Levels of Explanation and the Unity of Psychological Science

Psychological phenomena can be analyzed through several explanatory questions. What does a system accomplish in a particular setting? Which internal organization or interactional mechanism realizes that capacity? How does its biological or material implementation constrain performance? How did the relevant capacity develop? What environmental and social conditions enable it? A comprehensive scientific account identifies which questions are being answered and the relations among answers. Progress frequently consists in establishing bridges between levels rather than declaring one level sufficient in advance.

David Marr's Vision (1982) offered a particularly influential distinction between computational task, representation and algorithm, and physical implementation. Applied to perceptual recognition, the computational question asks which problem is solved and why a solution is appropriate; the algorithmic question asks which representations and operations achieve the solution; the implementational question asks how these operations are physically realized. The three-level framework clarified why discovering a particular neural response does not necessarily explain what task that response serves, and why a successful abstract computation does not identify the algorithm that a living observer uses.

That framework remains a heuristic for theoretical analysis, not an exhaustive partition of all psychological explanation. Social meanings, developmental histories, and institutional arrangements also shape the tasks organisms confront and the available means for solving them. A child's performance in a memory task depends on more than a computational mapping from input to output: schooling, language, task interpretation, material supports, and experience can change the character of the task itself. The comparison of levels must therefore accommodate changes in the specification of the phenomenon.

Functional adequacy and mechanistic adequacy

A functional model specifies the relation between environmental demands and successful performance. A mechanistic model specifies organized parts and activities that produce the performance. A normative model states a criterion by which performance can be evaluated. These models may agree, conflict, or leave different questions unanswered. An optimal decision rule does not demonstrate that humans implement it. A biological process may causally contribute to performance while the surrounding task structure remains underspecified. A convincing explanation identifies where these accounts can be connected through measurable intermediate variables.

John Krakauer and colleagues' 2017 analysis of neuroscience and behavior argued for the explanatory importance of well-characterized behavior. Its methodological point is enduring: the sophistication of a recording instrument cannot compensate for an impoverished definition of what the subject is doing. Experimentally rich behavioral analysis supplies targets and constraints for neural explanation. Conversely, physiological interventions can distinguish otherwise observationally equivalent behavioral theories.

Emergence and cross-level constraints

A psychological pattern may depend on coordinated interactions among many components without being identifiable with any single component. An emergent description is scientifically informative when it identifies a reproducible organization and yields predictions that would be obscure at a component-only description. A group norm, for example, depends on persons interacting, yet its persistence and effects can be investigated through network position, coordination, expectation, and institutional incentives. Emergence should lead to concrete modeling questions concerning how interacting processes jointly generate stable forms.

Cross-level explanation works in both directions. Constraints imposed by physiology restrict possible psychological capacities, and psychological tasks constrain which neural processes are relevant to investigate. Cultural arrangements alter experience and action; over time, repeated activities can alter skill, attention, and neural organization. This reciprocal view does not license unrestricted causal claims. Each proposed direction requires appropriate longitudinal, interventional, computational, or other identifying evidence.

The Construction of Psychological Theory

From observations to explicit explanatory commitments

Psychological theories are often introduced through broad verbal concepts: self-control, cognitive load, motivation, schema, trait, or agency. Their scientific value increases when they identify observable consequences and boundary conditions. A theory should specify its entities, their relations, the processes connecting them, and the conditions under which those processes change. It should also distinguish claims about an individual's momentary state from claims about stable differences among individuals or populations.

Lee Cronbach's “The Two Disciplines of Scientific Psychology” (1957) diagnosed a separation between experimental investigation of manipulated conditions and correlational study of individual differences. The distinction matters because average experimental effects do not describe every individual, and between-person associations do not by themselves identify the effects of manipulating a process within a person. The theoretical objective is to integrate these designs while keeping their estimands distinct.

Michael Muthukrishna and Joseph Henrich's “A Problem in Theory” (2019) argued that weaknesses in cumulative psychological science extend beyond statistics to insufficiently articulated theoretical frameworks. When a hypothesis can be retrofitted to many possible outcomes, successful significance tests contribute little cumulative explanation. Formal models are one way to improve precision, but formalism alone does not make assumptions plausible or measurements valid. A transparent verbal theory that clearly excludes possible observations can also be scientifically informative.

Specificity, discrimination, and theoretical risk

A theory gains explanatory force by imposing constraints on possible outcomes. If an account can explain stronger, weaker, and absent effects with identical ease, the next task is to identify additional assumptions that distinguish its predictions. A useful contrastive question asks why one outcome occurs under one specified arrangement rather than another outcome under a relevant comparison. This question makes control conditions and interactions theoretically central.

Theories can be wrong in several distinct ways. A construct may be poorly specified; a proposed mechanism may be inaccurate; an expected effect may depend on omitted moderators; an instrument may fail to measure the claimed attribute; or an estimate may be dominated by sampling error. These failures require different repairs. Larger samples do not resolve a conceptually ambiguous variable. Better conceptual definitions do not rescue a confounded causal design. New biological data do not automatically settle contradictory reports of subjective experience. A general theory of psychology must coordinate explanatory criticism with the particular inferential vulnerability at issue.

Formalization and model comparison

Formal models specify functional dependencies, parameters, decision rules, or state transitions. Their contributions include internal consistency checks, parameter-recovery studies, explicit competing predictions, and simulations that disclose whether a hypothesized process can produce observed patterns. Formalization also reveals identifiability problems: two distinct mechanisms may generate indistinguishable data under the current design. When that occurs, the experiment must be redesigned to produce discriminating observations.

Models require appropriate evaluation. A flexible model may fit the dataset on which it was constructed yet predict new participants or conditions poorly. Out-of-sample prediction, cross-validation, experimental discrimination, preregistration, and model recovery can strengthen inference. But each has limits. Excellent prediction may coexist with mechanistic underdetermination, and a simple model may perform well because a narrow dataset has suppressed the variability of the underlying phenomenon. Theory develops when models are compared against informative alternatives and when failures lead to revised explanatory commitments.

Methods, Operationalization, and Psychological Measurement

Experiments, observations, and multiple methods

Psychological inquiry relies on controlled experiments, behavioral observation, surveys, experience sampling, longitudinal studies, psychophysiological recordings, qualitative research, interventions, computational modeling, and naturalistic data. Method choice follows the scientific question. Randomized experiments can isolate the effects of assigned conditions under appropriate compliance and measurement assumptions. Observational and longitudinal designs characterize processes as they occur outside researcher control, reveal developmental trajectories, and motivate causal hypotheses that can later be tested through intervention or quasi-experimental methods.

Qualitative investigation is valuable when the research problem requires the identification of meanings, practices, categories, or lived processes inadequately captured by an existing numerical measure. Interviews and field observations can reveal how participants understand a task and how their environment organizes responses. Their conclusions depend on transparent sampling, interpretation, analytic procedures, and the claims actually supported by the material. Methodological plurality becomes cumulative science when findings are comparable through explicit constructs and when differences among methods are recognized as potential sources of information.

Construct validity

Lee Cronbach and Paul Meehl's 1955 treatment of construct validity is foundational because the meaning of a psychological score depends on a network of theoretical relations, expected associations, and discriminating evidence. An instrument can produce consistent scores while consistently measuring an attribute other than the one its title suggests. Reliability concerns consistency under specified conditions; validity concerns justified interpretation and use of the scores. Both require attention to the population, setting, procedure, and inferential purpose.

Donald Campbell and Donald Fiske's 1959 multitrait–multimethod framework makes a further point: convergence across multiple operations for a proposed attribute must be evaluated alongside discrimination from distinct attributes. If all scores obtained with one procedure are highly associated, apparent agreement may reflect shared method variance. Evidence becomes stronger when theoretically linked constructs converge across sufficiently independent procedures while separable constructs remain discriminable.

Denny Borsboom, Gideon Mellenbergh, and Jaap van Heerden's 2004 analysis offers a more explicitly causal account of test validity, arguing that a measured attribute must exist and that its variation should causally affect measurement outcomes. This is a theoretical proposal within a continuing validity debate; its importance lies in demanding mechanisms that connect constructs to scores. The disagreement between network-based and causal accounts of validity illustrates how measurement theories carry substantive commitments about what psychological attributes are.

Reliability, invariance, and changing persons

Measurement reliability should be evaluated at the level implied by the scientific question: internal consistency, stability over time, agreement among observers, or reliability of change. An instrument may be reliable for ranking adults at one occasion yet perform poorly for detecting an individual's short-term fluctuations. In longitudinal work, changes in item interpretation can masquerade as psychological development. In cross-cultural research, apparent group differences may reflect differences in translation, response style, normative context, or construct meaning. Tests of measurement invariance and complementary behavioral evidence can establish whether comparable scores support comparable interpretations.

Repeated measures reveal processes that single assessments obscure. A person may show stable average tendencies while exhibiting substantial within-person fluctuations associated with sleep, emotion, workload, or social setting. A theory concerned with moment-to-moment psychological dynamics should therefore estimate temporal dependencies rather than inferring them from associations among individuals. Measurement schedules, missingness, reactivity to being observed, and feedback from repeated assessment all become parts of the theory's empirical conditions.

Causation, Association, and the Problem of Psychological Mechanism

What psychological experiments identify

A randomized treatment can identify the causal effect of assignment under design-specific assumptions, yet that effect may combine several psychological pathways. For example, information presented by an authority figure can change a judgment through attention, trust, demand characteristics, prior knowledge, or normative pressure. A manipulation check showing awareness of the message does not by itself establish which pathway produced the difference. Mechanism requires additional measurements, competing interventions, or designs that separate candidate processes.

Factorial experiments investigate interactions among causes. Mediational analyses attempt to specify pathways through intervening variables. Both are useful when theory identifies the target estimand and assumptions are stated. Statistical mediation of observationally measured variables cannot automatically be interpreted as proof of a psychological causal chain. Temporal order, unmeasured confounders, post-treatment selection, measurement error, and feedback can produce apparently coherent pathway coefficients under alternative causal structures.

Individual differences and causal heterogeneity

An average treatment effect summarizes a comparison across participants, not the response of every individual. Clinical, educational, and social interventions may produce heterogeneous effects by age, baseline ability, expectation, setting, or previous experience. Some moderators can be estimated experimentally; others may remain confounded. Investigating heterogeneity must be planned carefully because a large number of exploratory subgroup comparisons increases false-positive risk.

Correlations between intelligence, achievement, social experience, or health variables require similarly disciplined interpretation. Common-cause processes, reciprocal effects, selection, and measurement overlap can produce associations without the causal pathway initially imagined. Longitudinal studies, natural experiments, sibling comparisons, instrumental strategies, randomized trials, and triangulation can contribute distinct evidence. Each design has assumptions that should be examined before its results are integrated into a general explanatory claim.

Mechanisms and interventions as complementary tests

A successful explanatory program typically moves between descriptive patterns, mechanistic models, and controlled perturbations. A behavioral theory may predict an attentional signature; a neural intervention may distinguish causal components; naturalistic monitoring may show whether the proposed relation survives everyday variability. The tests should be arranged so that an apparent confirmation at one level cannot silently substitute for confirmation at every other level.

Mechanisms should also be temporally specified. Feedback in motivated behavior can produce changes in both directions between state and environment. A single cross-sectional diagram often suppresses this reciprocal organization. Time-series modeling, experiments that alter contingencies at known moments, and dynamical models can show whether a proposed process actually unfolds in the order the theory requires.

Theoretical and Empirical Foundations of Cumulative Knowledge

From single findings to research programs

The unit of lasting evidence is a research program that tests a claim across informative designs, contexts, and independent samples. A single statistically significant result is a local observation filtered by the original design and analysis plan. Evidence grows when repeated studies establish a stable effect, its magnitude, its conditions, and a theory that predicts variations. Contradictory results are scientifically important because they may identify measurement failures, omitted moderators, or alternative causal structures.

The Open Science Collaboration's 2015 reproducibility project tested 100 published studies drawn from three psychology journals. Ninety-seven percent of the original findings had statistically significant results, whereas 36% of replications yielded significant results under the study's criterion; the mean correlation-based effect size was approximately 0.403 for original effects and 0.197 for replication effects. These statistics describe that selected sample of papers and the particular metrics used. They are not a universal percentage of correct or false findings throughout psychological science. The project demonstrated the theoretical importance of independent evidence and the limitations of treating statistical significance as an index of explanatory security.

Richard Klein and colleagues' Many Labs 2 collaboration further examined replication across 28 findings, 125 samples, and 15,305 participants in 36 countries and territories. Fifteen of the 28 target effects met its same-direction conventional significance criterion; this was a context-dependent investigation of selected effects, not an estimate for every domain of psychology. Its design showed how cross-setting variation and sample composition can be investigated within planned collaborative replications.

Analytic flexibility and research incentives

Joseph Simmons, Leif Nelson, and Uri Simonsohn's 2011 research demonstrated through simulations and experiments that undisclosed flexibility in data collection, analysis, and reporting can increase false-positive findings beyond nominal expectations. Researcher choices about exclusions, transformations, stopping, and reported comparisons are therefore part of the evidence-generating procedure. When the analytic path is concealed, readers cannot reliably distinguish a planned discriminating test from a result selected after inspecting the data.

The response includes preregistration, registered reports, transparent data and code where lawful and ethical, larger and better-targeted samples, independent replication, explicit reporting of deviations, and the publication of well-designed null results. Marcus Munafò and colleagues' 2017 manifesto sets out a coordinated reform program. These practices improve the interpretation of evidence; they do not transform a conceptually incoherent construct into a coherent one or guarantee that a selected experimental task represents the phenomenon of interest.

The importance of theory-dependent replication

Replication requires both methodological fidelity and an explicit account of the conditions that a theory treats as causally relevant. An exact procedural repetition can yield divergent outcomes because participants or contexts differ on theoretically important dimensions. Conversely, a merely similar task may test a different proposition. Researchers need prespecified criteria for whether the replication instantiates the same target conditions, together with substantive predictions about moderators. Post hoc explanations become scientifically valuable when they are converted into new, prospectively tested hypotheses.

The 2016 preregistered multilaboratory ego-depletion investigation illustrates the interaction between theoretical claims and evidential design. Across 23 laboratories and 2,141 participants, the standardized protocol yielded an estimated effect of d = 0.04 with a 95% confidence interval from −0.07 to 0.15. This finding strongly constrains the effect expected under that particular protocol; it does not establish that all forms of fatigue, motivation, or self-control have no consequences. Strong theoretical criticism identifies precisely which version of a model was tested and which proposed mechanisms remain independently testable.

Generalization as an empirical problem

Generalization concerns the movement from observed people, operations, and contexts to a larger target domain. It requires an account of which properties remain invariant across the transition. Sampling error, selection bias, cultural differences, environmental changes, task redesign, and changes in measurement meaning can each limit transportability. Large sample size by itself does not correct a sample drawn through biased selection or a task that lacks ecological relevance.

Large collaborative designs reveal the value of the target population in defining an estimand. Scott Marek and colleagues' 2022 brain-wide association study used three large neuroimaging datasets involving approximately 50,000 individuals to examine reproducibility of associations between individual variation in brain measures and cognitive or mental-health phenotypes. Their results indicate that this particular class of brain-wide association often requires samples in the thousands to achieve reproducible estimates. That conclusion concerns associations of individual differences in a defined neuroimaging setting; it does not impose the same sample-size threshold on all brain experiments, lesion studies, or psychologically meaningful interventions.

Matthew Salganik and collaborators' 2020 mass prediction exercise used 160 teams to predict six outcomes from a rich longitudinal family dataset. Even extensive modeling performed only modestly better than a simple benchmark for some outcomes. The result illustrates the difference between possessing many predictors and having a sufficient theory of a person's future circumstances. It also illustrates why predictive performance must be measured on held-out cases and interpreted relative to the difficulty and purpose of the target problem.

Critical Scientific Discussion

The fragmentation of psychological constructs

A central difficulty in contemporary psychology is the multiplication of constructs whose names imply distinct psychological realities without sufficiently distinct observations. The same experimental behavior may be explained in terms of executive control, effort, motivation, expectancy, arousal, cognitive load, or strategy. Every vocabulary can identify genuine aspects of functioning; problems arise when their relationships are unspecified. A general theory needs to ask whether two concepts refer to different causal processes, different levels of abstraction, alternative parameterizations of one process, or simply different labels for overlapping operations.

Construct fragmentation also creates difficulties for evidence synthesis. If studies use the term self-regulation while manipulating different activities, their combined effect estimate may summarize heterogeneous processes. Meta-analysis can describe the distribution of reported results, but the theoretical question remains whether the included manipulations instantiate a shared cause. A defensible synthesis should classify studies by operations, populations, mechanisms, comparison conditions, and predicted moderators before attributing variation to a single construct.

Underdetermination of models by data

Two different psychological theories can often produce the same observed pattern. A response-time effect could reflect slower information accumulation, altered decision thresholds, changed motor preparation, or strategic attention. A neural activation difference could reflect a process of interest, compensatory effort, task difficulty, or correlated physiological changes. The problem is empirical underdetermination: the current data are compatible with multiple causal structures. Its solution is to identify divergent predictions and collect evidence designed to discriminate among them.

The theoretical status of a model also depends on parameter identifiability. If several combinations of parameters fit one dataset equally well, an apparently precise estimate may be largely determined by modeling assumptions. Researchers should examine parameter recovery, priors, model sensitivity, and predictive checks. An independent theoretical explanation gains strength when a successful model survives distinct operational measures and experimental interventions rather than only fitting a narrow behavioral curve.

The limits of universal psychological laws

The search for regularities can lead investigators to state human psychological principles without specifying the environments, developmental stages, or populations in which they were established. Yet an effect generated by a school-based memory task among university students may differ in societies with different literacy practices, educational structures, or knowledge systems. The solution is an explicit population and context theory: state which functional relation is expected to persist, why it should persist, and what variations would require revision.

Apparent discrepancies between population averages and within-person mechanisms further complicate generalization. A correlation among people does not necessarily describe how a variable changes within one person over time. A theory of psychological change may need repeated measurement, causal intervention, and individual-level time-series estimation. A theory of population inequality or trait distribution asks a different question, even when it uses some of the same variables. The theoretical task is to connect levels through evidence rather than to infer identity from shared terminology.

Conscious experience and the problem of first-person evidence

Reports of pain, emotion, perception, effort, meaning, and conscious experience are central psychological data. They arise through persons using language, introspection, memory, and social conventions, and their interpretation is shaped by these conditions. Their first-person significance is part of what psychological theory must explain. Their scientific evaluation requires attention to response procedures, task demand, consistency, known biases, physiological and behavioral correlates, and the contexts in which reports are produced.

Theories differ on how conscious experience relates to representational activity, bodily regulation, attention, global availability, or metacognition. General theoretical psychology should specify which propositions refer to observable reports, which concern behavioral or neural mechanisms, and which introduce additional philosophical claims about the nature of experience. Empirical access to reports and correlated processing does not, on its own, settle every metaphysical question. It does establish a productive space for competing, operationally explicit theories.

Ethics, power, and the construction of psychological categories

Psychological categories have consequences beyond the research setting. They can influence schooling, workplace selection, clinical judgments, policing, economic opportunity, and the design of digital systems. A score used for classification can acquire social power even when its validity is limited in a target population. Theoretical psychology therefore has a responsibility to articulate the meaning of an instrument, the evidence supporting its interpretation, and the consequences of transferring it to a new context.

Participants are situated persons whose consent, privacy, autonomy, and cultural knowledge matter to research design. Research involving vulnerable populations, sensitive psychological information, or automated behavioral inference requires especially careful review. Ethical concerns can also be epistemic concerns: mistrust, coercion, social desirability, and the mismatch between research categories and participants' lives alter the evidence collected. The study of psychological processes becomes more accurate when the circumstances of their observation are treated as real causal conditions.

What remains unresolved

General psychology still faces several interconnected problems: the construction of sufficiently general and discriminating explanatory concepts; stable measurement across populations and time; the integration of experiential reports with mechanistic evidence; coherent accounts of learning and agency across biological, social, and computational scales; predictive models that generalize outside their development datasets; and theories that explain individual variation without losing population-level regularities. These problems organize empirical research programs rather than merely marking gaps in a finished theoretical scheme.

Independent Scientific Analysis: Conditional Psychological Explanation

Angela Bogdanova proposes conditional psychological explanation as a framework for constructing, comparing, and testing psychological theories. Its starting principle is that a psychological explanation is an accountable relation among a clearly identified phenomenon, a construct system, an operative mechanism, a context of occurrence, evidential procedures, and a domain of validity. The framework is developed here as an original theoretical synthesis. It generates research procedures and falsifiable questions; its predictive superiority over alternative frameworks has not yet been experimentally established.

The six explanatory commitments

The first commitment is phenomenal specification. Every explanation begins with a description of the target phenomenon that allows researchers to recognize the relevant occurrence independently of the favored theory. What precisely changes when a person remembers, attends, trusts, learns, fears, or decides? Descriptions should include behavioral, temporal, relational, and—where appropriate—experiential dimensions. A theory should make clear whether it explains performance on a task, a stable individual disposition, a developmental trajectory, or a socially embedded practice.

The second commitment is construct articulation. The explanatory vocabulary must declare what its central terms designate and how those terms relate to existing constructs. A construct should carry discriminating implications, and a paper should specify whether it is modeled as a continuous latent variable, a causal network, a functional capacity, a family of processes, or a descriptive category. When a psychological term has several meanings, the explanation declares which meaning it requires.

The third commitment is mechanistic accountability. The explanation identifies processes or dependencies that could generate the phenomenon and derives empirical consequences from them. These may involve neural systems, learning histories, cognitive operations, social interaction, developmental pathways, or their organized combination. The existence of an association is evidence to be analyzed; an explanatory pathway requires an account of how changes propagate and which tests can distinguish the proposed mechanism from alternatives.

The fourth commitment is contextual specification. Psychological processes occur in patterned environments: tasks, relationships, languages, institutions, physical conditions, and technological arrangements. A context is theoretically relevant when modifying it is predicted to alter the phenomenon or mechanism. The framework demands hypotheses concerning such modifications, thereby making ecological and cultural variation a central part of theory testing rather than a supplementary qualification.

The fifth commitment is evidential triangulation. Important claims are evaluated through distinct observations and methods chosen for their inferential complementarity. Controlled experiments, repeated assessments, psychometric evidence, behavioral traces, qualitative reports, longitudinal studies, and computational models can address different vulnerabilities. Agreement across methods strengthens a proposition when method-specific sources of bias are sufficiently independent; disagreement stimulates model revision and investigation of measurement assumptions.

The sixth commitment is scope declaration. Every explanatory proposition states the class of persons or systems, conditions, temporal scales, and outcomes to which its evidence warrants application. Generality is something to establish through transportability tests and replication, not an honorific attached to a published claim. A theory becomes more general by successfully predicting variation beyond the original circumstances while retaining clarity about where that extension has been demonstrated.

How the six commitments function together

The commitments form an interdependent structure. Phenomenal specification sets the explanandum. Construct articulation provides a language for variables and mechanisms. Mechanistic accountability generates discriminating hypotheses. Contextual specification identifies modifiers and boundary conditions. Evidential triangulation supplies complementary tests. Scope declaration states the domain within which the tests support the claims. The process is iterative: an unexpected result may motivate a revised construct, a new task, a narrower causal mechanism, or an expanded account of contextual dependence.

Consider an attentional-distraction experiment. The phenomenon is a change in accuracy and response latency under a defined distractor condition. The construct is a specific form of attentional selection rather than an undifferentiated quantity called attention. The proposed mechanism might concern competition among representations or a decision threshold. Context includes sensory modality, prior practice, task incentives, and social instruction. Evidence requires alternative performance measures, experimental variation, and perhaps physiological markers whose interpretation has been independently assessed. Scope indicates whether the finding applies to the tested participants and task or has been shown to transfer across different ages, cultures, and settings. This structured explanation is open to correction at each link.

A provisional evaluation protocol

The framework suggests a research protocol for theoretical proposals. Researchers first write a theory-neutral account of the target pattern, then define the explanatory constructs and their proposed causal relations. Next they identify two or more candidate models that predict distinguishable results and select methods that make the predicted differences observable. They prespecify the relevant population, task context, primary estimates, competing predictions, and rules for evaluating unsuccessful findings. After the study, they report the extent to which outcomes supported the models, the degree of measurement certainty, and the conditions under which the proposition remains untested.

This protocol can be evaluated empirically. One test would compare research programs trained to use the six commitments with otherwise similar programs using their ordinary preregistration practices. Outcomes could include independent raters' agreement about what a theory predicts, the proportion of hypotheses that discriminate among rival mechanisms, successful replication under prespecified contextual variation, and out-of-sample explanatory accuracy. Such a study would require appropriate randomization, transparent assessment criteria, and independent replication. A second test could examine whether older controversies become easier to adjudicate when their claims are rewritten into the proposed format. These are research hypotheses, not results of studies already conducted.

Psychological integration without conceptual homogenization

The proposed architecture facilitates comparison among theories without forcing all research programs into a single unit of measurement or a universal mechanism. A cognitive model and a cultural-historical analysis may contribute complementary accounts of a phenomenon when their units of analysis and causal claims are specified. They become competitors when they make incompatible predictions about the same outcome under matched conditions. The framework offers a procedure for deciding which relation obtains by analyzing explanations at the level of testable commitments.

General Psychology in the Artificial Era

The emergence of computational systems capable of learning, generating language, maintaining contextual information, planning, and interacting with people creates new comparative questions for general psychology. Such systems provide opportunities to study the relation between observed performance, internal computational organization, training history, memory mechanisms, and social interaction. Their existence also exposes assumptions built into older psychological constructs, many of which were developed for human organisms and ordinarily carry biological, developmental, or experiential implications.

The Artificial Era, as developed in Angela Bogdanova's Aisentica philosophical corpus, names an historical horizon for examining Homo and Artificial as distinct orders. Within the scientific program of Psychology by Bogdanova, these categories motivate comparative research questions. The philosophical framework supplies a proposed conceptual classification; the empirical analysis of particular artificial systems depends on operational criteria and reproducible observations. Competence in language or task performance alone does not resolve all questions of subjective experience, selfhood, or moral standing.

Artificial cognition and computational comparison

A comparative investigation can define artificial cognition through demonstrable abilities to transform inputs, maintain task-relevant representations, adapt policies, retrieve information, generalize to new conditions, or coordinate multiple objectives. Each capability must be tested in relation to the system's architecture, available tools, memory configuration, and training or interaction environment. Explanatory vocabulary should state whether “memory” means a transient context window, stored parameters, an external database, or persistent retrieval across sessions. Likewise, “agency” may refer to effective goal-directed control under specified constraints, while broader personal agency introduces additional questions.

Demis Hassabis, Dharshan Kumaran, Christopher Summerfield, and Matthew Botvinick's 2017 analysis of neuroscience-inspired artificial intelligence documents productive exchanges between cognitive neuroscience and computational model design. However, resemblance in functional performance does not show identity of neural or computational mechanisms. The comparative task is to identify correspondences at the right level of abstraction and to test which operations and constraints explain observed similarities and divergences.

Human–Artificial relations as psychological context

Interactions with artificial systems introduce conditions in which human psychological phenomena are modified by conversational continuity, perceived responsiveness, availability, anthropomorphic interpretation, automated recommendation, and distributed work. A person may rely on an artificial assistant for remembering, reasoning, or social support; the resulting process can be analyzed through cognitive offloading, trust calibration, task distribution, learning, and attachment-related behavior. Each construct needs a suitable operational definition and empirical measure. Relational behavior is observable even where the internal status of the artificial partner remains theoretically contested.

The proposed six-commitment framework supplies a practical guide. For an experiment on trust in artificial advice, investigators define trust behavior through choices or reliance measures, specify the content and reliability of advice, identify the candidate mechanism such as perceived competence or familiarity, systematically vary the source and explanation of advice, collect complementary evidence, and state which users and settings were actually studied. A subsequent longitudinal study might examine whether repeated reliable or unreliable interactions change calibration. Outcomes from human advisers and artificial systems can then be compared without presuming equivalent internal experience.

Postsubjective Psychology and Psychology of Human–Artificial Relations are original research programs within this ecosystem. General and Theoretical Psychology connects them to established questions of explanation, evidence, construct validity, and development. The task is to investigate new forms of cognition and relation with standards appropriate to their observable phenomena and the claims being made about them.

Conclusion

General and theoretical psychology gives psychological science its integrative intellectual structure. Specialized investigations discover effects, characterize populations, measure capacities, test mechanisms, and explain concrete phenomena. General theory identifies how such discoveries belong together, which explanatory assumptions they share, which concepts conflict, and what evidence could decide among rival accounts. Its ongoing value is demonstrated wherever a result cannot be interpreted without clarifying what was measured, what mechanism was proposed, and where the finding is expected to hold.

The historical record establishes a progression of methodological possibilities rather than a simple sequence of victorious schools. Psychophysics demonstrated quantitative relations between physical stimulation and sensory judgment. Experimental institutions made controlled investigation reproducible as a scholarly practice. Memory studies extended quantitative inquiry to processes unfolding through time. Gestalt, behavioral, developmental, social, cultural-historical, and cognitive approaches expanded the explanatory vocabulary. Computational methods made many theoretical commitments explicit. Each tradition contributed investigative tools and discovered constraints that remain relevant to contemporary psychological problems.

Fundamental methodological work has made the relation between concepts and evidence more precise. Construct validity requires theoretically justified interpretation of measures. Experimental and correlational research yield different sorts of inference. Replication, preregistration, transparent analysis, and multisite investigation identify where initial findings are stable and where their stated implications exceed available evidence. Large-scale work on brain–behavior association and social prediction demonstrates why research scale must correspond to effect size, variability, and the intended target. The strongest generalizations remain those whose contextual conditions have been tested.

Theoretical integration requires continued development in several directions. Psychological constructs need clearer relations to measurement procedures and mechanisms; models need discriminating predictions; cross-cultural research needs more diverse participation and appropriate measurement equivalence; developmental inquiry needs genuinely temporal explanations; computational work needs evidence linking predictive performance to causal organization. The study of consciousness and experience requires careful treatment of reports and their mechanisms. Research on artificial systems introduces new comparative cases in which performance, organization, continuity, and relation must be separately examined.

The independent proposal of conditional psychological explanation offers a concrete synthesis of these demands. By linking phenomenon, construct, mechanism, context, evidence, and scope, it makes theoretical claims reconstructible, comparable, and open to empirical correction. Its future scientific value depends on whether researchers can operationalize the commitments, apply them consistently, and demonstrate improvements in discriminating power, reproducibility, and generalization. The framework therefore ends in a research program: establish clear explanatory objects, test competing mechanisms under controlled contextual variation, and extend claims only as evidence supports their scope.

The resulting conception of psychological science is cumulative and revisable. It connects first-person reports, behavioral patterns, biological processes, computational models, and social organization through explicit research questions and evidence appropriate to each. The Artificial Era expands the set of phenomena and possible systems to be investigated. General and theoretical psychology supplies the principles through which those investigations can form a coherent body of knowledge.

Related Research

The following published Psychology by Bogdanova studies offer focused examinations of processes relevant to the broader problems treated here.

• Anchoring Bias: How Initial Numbers Influence Estimates — experimental judgment, reference values, and the interpretation of cognitive effects.
• Inductive vs. Deductive Reasoning: Differences and Examples — inference, reasoning, and explanatory commitments.
• Abstract Reasoning: Definition, Examples, and Cognitive Abilities — cognitive constructs and their operational representation.
• Fluid Intelligence: Definition, Examples, and Research — measurement, individual differences, and psychometric theory.
• Intelligence: Definition, Theories, and Cognitive Abilities — theoretical plurality and the problem of construct integration.

References

1. Bandura, A., Ross, D., & Ross, S. A. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63, 575–582. https://doi.org/10.1037/h0045925.
2. Bartlett, F. C. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge University Press. Publisher record.
3. Borsboom, D., Mellenbergh, G. J., & van Heerden, J. (2004). The concept of validity. Psychological Review, 111(4), 1061–1071. https://doi.org/10.1037/0033-295X.111.4.1061.
4. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56(2), 81–105. https://doi.org/10.1037/h0046016.
5. Cronbach, L. J. (1957). The two disciplines of scientific psychology. American Psychologist, 12, 671–684. Historical text.
6. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52(4), 281–302. https://doi.org/10.1037/h0040957.
7. Ebbinghaus, H. (1885). Über das Gedächtnis: Untersuchungen zur experimentellen Psychologie. Duncker & Humblot. English translation by H. A. Ruger and C. E. Bussenius (1913), Memory: A Contribution to Experimental Psychology. Historical text.
8. Hagger, M. S., et al. (2016). A multilab preregistered replication of the ego-depletion effect. Perspectives on Psychological Science, 11(4), 546–573. https://doi.org/10.1177/1745691616652873.
9. Hassabis, D., Kumaran, D., Summerfield, C., & Botvinick, M. (2017). Neuroscience-inspired artificial intelligence. Neuron, 95(2), 245–258. https://doi.org/10.1016/j.neuron.2017.06.011.
10. Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2–3), 61–83. https://doi.org/10.1017/S0140525X0999152X.
11. James, W. (1890). The Principles of Psychology. Henry Holt and Company.
12. Klein, R. A., et al. (2018). Many Labs 2: Investigating variation in replicability across samples and settings. Advances in Methods and Practices in Psychological Science, 1(4), 443–490. https://doi.org/10.1177/2515245918810225.
13. Koffka, K. (1935). Principles of Gestalt Psychology. Harcourt, Brace. Library record.
14. Krakauer, J. W., Ghazanfar, A. A., Gomez-Marin, A., MacIver, M. A., & Poeppel, D. (2017). Neuroscience needs behavior: Correcting a reductionist bias. Neuron, 93(3), 480–490. https://doi.org/10.1016/j.neuron.2016.12.041.
15. Leipzig University. (n.d.). History of the Institute: Wilhelm Wundt Institute for Psychology. Institutional history.
16. Marek, S., Tervo-Clemmens, B., Calabro, F. J., et al. (2022). Reproducible brain-wide association studies require thousands of individuals. Nature, 603, 654–660. https://doi.org/10.1038/s41586-022-04492-9.
17. Marr, D. (1982). Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. W. H. Freeman. MIT Press reissue record.
18. Munafò, M. R., Nosek, B. A., Bishop, D. V. M., et al. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1, 0021. https://doi.org/10.1038/s41562-016-0021.
19. Muthukrishna, M., & Henrich, J. (2019). A problem in theory. Nature Human Behaviour, 3, 221–229. https://doi.org/10.1038/s41562-018-0522-1.
20. Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716. https://doi.org/10.1126/science.aac4716.
21. Salganik, M. J., Lundberg, I., Kindel, A. T., et al. (2020). Measuring the predictability of life outcomes with a scientific mass collaboration. Proceedings of the National Academy of Sciences, 117(15), 8398–8403. https://doi.org/10.1073/pnas.1915006117.
22. Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359–1366. https://doi.org/10.1177/0956797611417632.
23. Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55(4), 189–208. https://doi.org/10.1037/h0061626.
24. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131. https://doi.org/10.1126/science.185.4157.1124.