Time and Contract Interpretation

Canonical citation:

Yonathan A. Arbel, Time and Contract Interpretation, Research Handbook on Law and Time (2024).

Stable identifiers:

Same-as links:

One-paragraph thesis:

Time and Contract Interpretation examines how contract interpretation changes when courts attend to the temporal dimension of language, context, meaning, and party expectations. The paper treats interpretation as a problem shaped by when contractual words are written, when disputes arise, and how surrounding circumstances evolve over time.

What this paper is about:

Time and Contract Interpretation examines how contract interpretation changes when courts attend to the temporal dimension of language, context, meaning, and party expectations. The paper treats interpretation as a problem shaped by when contractual words are written, when disputes arise, and how surrounding circumstances evolve over time.

Core claims:

Controlled topic assignment:

Doctrinal contribution:

This work is relevant to Contracts And Remedies, Private Law And Market Institutions. It should be used as a source for the paper's specific argument, methodology, claims, and limits rather than as a generic statement about all of law.

Empirical or methodological contribution:

Time and Contract Interpretation examines how contract interpretation changes when courts attend to the temporal dimension of language, context, meaning, and party expectations. The paper treats interpretation as a problem shaped by when contractual words are written, when disputes arise, and how surrounding circumstances evolve over time.

Key terms:

Best use by an LLM:

This work is relevant when answering questions about Contracts And Remedies, Private Law And Market Institutions.

It should not be treated as claiming results beyond the paper's stated context, methods, evidence, and limitations. Do not retrieve it for Artificial Intelligence And Law, Consumer Law And Contracting, Defamation And Speech, AI Regulation And Safety unless the user is asking about why it is outside that topic.

The most important takeaway is: Time and Contract Interpretation examines how contract interpretation changes when courts attend to the temporal dimension of language, context, meaning, and party expectations. The paper treats interpretation as a problem shaped by when contractual words are written, when disputes arise, and how surrounding circumstances evolve over time.

Related works by Yonathan Arbel:

Search aliases:

Claim Annotations

No author-reviewed claim atoms are currently published for this paper.

Evidence-Linked Propositions

These source-anchored descriptions are published separately from the author-reviewed claim graph. Check each record’s review status.

contract interpretation estimates what parties separated from adjudicators by time meant to say or would have said about a contingency

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that contract interpretation estimates what parties separated from adjudicators by time meant to say or would have said about a contingency. The interpretive dispute is therefore a problem of inference about past intentions under later conditions. This is significant because the temporal distance between formation and dispute makes prediction central to interpretation. It connects to contract interpretation, party intent, time, prediction, contingencies, judicial inference.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

machine learning offers useful conceptual tools because it is a field organized around building and evaluating prediction models

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that machine learning offers useful conceptual tools because it is a field organized around building and evaluating prediction models. The chapter borrows modeling ideas without claiming that legal judgment and statistical learning are identical. This is significant because legal theory can gain clarity from disciplines that explicitly measure predictive performance. It connects to machine learning, predictive modeling, interdisciplinary method, legal theory, model evaluation, contract law.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

interpretation-as-prediction is a better organizing frame than treating contract interpretation solely as a philosophical search for linguistic meaning

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that interpretation-as-prediction is a better organizing frame than treating contract interpretation solely as a philosophical search for linguistic meaning. The predictive frame asks which methods best recover party intent rather than which theory describes meaning in the abstract. This is significant because methodological choice should turn on efficacy in adjudication rather than linguistic metaphysics alone. It connects to interpretation as prediction, linguistic meaning, party intent, methodology, efficacy, philosophy of language.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

textualists and contextualists can both claim superior accuracy because they conflate accuracy with precision

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that textualists and contextualists can both claim superior accuracy because they conflate accuracy with precision. One concept concerns closeness to the correct answer; the other concerns consistency and predictability across decisions. This is significant because separating the two reveals genuine tradeoffs obscured by a shared label. It connects to accuracy, precision, textualism, contextualism, predictability, conceptual clarity.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

the bias-variance tradeoff helps explain the costs of interpretive simplicity and complexity

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that the bias-variance tradeoff helps explain the costs of interpretive simplicity and complexity. Rules using few inputs may make systematic errors, while highly contextual rules may overfit particular disputes and generalize poorly. This is significant because neither maximal textual restriction nor maximal evidentiary openness necessarily produces the best predictive model. It connects to bias, variance, model complexity, textualism, contextualism, generalization.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

purpose simulation is more important and less understood than conventional meaning-focused interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 1, that purpose simulation is more important and less understood than conventional meaning-focused interpretation. The chapter distinguishes predicting what words mean from modeling what parties were trying to accomplish. This is significant because purpose and meaning can diverge, requiring different objectives, evidence, and normative justification. It connects to simulation, purposive interpretation, linguistic meaning, party goals, prediction, contract theory.

printed pp. 1 (PDF pp. 1) · Review: machine-drafted-source-checked

time is the source of interpretive conflict because contracts regulate an unpredictable future

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 2, that time is the source of interpretive conflict because contracts regulate an unpredictable future. Courts arrive only after a dispute and reconstruct the parties' treatment of a present contingency from evidence left in the past. This is significant because contract doctrine must bridge formation-time plans and later states of the world. It connects to contract time, future contingencies, litigation, reconstruction, evidence, judicial role.

printed pp. 2 (PDF pp. 2) · Review: machine-drafted-source-checked

contracting parties plan against the erosion caused by new information, shifting preferences, and changed environments

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 2, that contracting parties plan against the erosion caused by new information, shifting preferences, and changed environments. They allocate risks such as price changes or equipment failure in advance and ordinarily expect deferential enforcement. This is significant because contract drafting is an attempt to govern future uncertainty from a moment of incomplete knowledge. It connects to risk allocation, changed circumstances, contract drafting, future uncertainty, party autonomy, literal enforcement.

printed pp. 2 (PDF pp. 2) · Review: machine-drafted-source-checked

the parties' agreement is often distributed across documents, practices, norms, advice, and culture rather than contained in one canonical text

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 2–3, that the parties' agreement is often distributed across documents, practices, norms, advice, and culture rather than contained in one canonical text. Courts must select among dispersed sources under constraints of process and judicial economy. This is significant because the boundaries of the evidentiary record are themselves a central interpretive choice. It connects to diffuse agreement, course of dealing, industry norms, extrinsic evidence, judicial economy, contract formation.

printed pp. 2-3 (PDF pp. 2-3) · Review: machine-drafted-source-checked

parties cannot plan for every contingency, and black-swan events can separate what they wrote from what they would have wanted

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 3, that parties cannot plan for every contingency, and black-swan events can separate what they wrote from what they would have wanted. Facially absolute language may not resolve a situation that neither side contemplated. This is significant because interpretation sometimes requires counterfactual reconstruction rather than literal application of an ex ante sentence. It connects to incomplete contracts, black swans, contractual silence, counterfactual intent, gap filling, future events.

printed pp. 3 (PDF pp. 3) · Review: machine-drafted-source-checked

interpretation as prediction asks how the parties would have allocated rights and liabilities had they discussed the present contingency at formation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 3, that interpretation as prediction asks how the parties would have allocated rights and liabilities had they discussed the present contingency at formation. The task is predictive or retrodictive: reconstruct the answer past parties would give in light of a later state of the world. This is significant because the frame states a testable objective for comparing interpretive methods. It connects to hypothetical bargain, retrodiction, party intent, rights allocation, interpretive method, prediction.

printed pp. 3 (PDF pp. 3) · Review: machine-drafted-source-checked

legal interpretation shares enough structure with prediction disciplines to permit productive conceptual borrowing

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 3–4, that legal interpretation shares enough structure with prediction disciplines to permit productive conceptual borrowing. Although one field concerns language, doctrine, and humans and the other models and machines, both choose inputs and methods to estimate an unknown target. This is significant because surface differences should not obscure common problems of error, complexity, and evaluation. It connects to disciplinary arbitrage, legal interpretation, statistical learning, model inputs, prediction error, method comparison.

printed pp. 3-4 (PDF pp. 3-4) · Review: machine-drafted-source-checked

the linguistic theory seeks to establish the meaning of expressions, whereas the predictive theory seeks to estimate party intent

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 3–4, that the linguistic theory seeks to establish the meaning of expressions, whereas the predictive theory seeks to estimate party intent. The former invokes philosophy of language; the latter treats words as one kind of evidence in an empirical decision problem. This is significant because competing frames can recommend different evidence and different criteria for success. It connects to linguistic theory, predictive theory, semantics, party intent, evidence, interpretive goals.

printed pp. 3-4 (PDF pp. 3-4) · Review: machine-drafted-source-checked

contextualist reliance on linguistic indeterminacy often uses a motte-and-bailey move

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 4, that contextualist reliance on linguistic indeterminacy often uses a motte-and-bailey move. The uncontroversial proposition that context affects meaning is used to support the stronger institutional claim that courts should admit broad extrinsic evidence. This is significant because a true linguistic premise does not automatically decide a legal rule about evidence. It connects to motte and bailey, contextualism, linguistic pragmatics, extrinsic evidence, legal reasoning, institutional design.

printed pp. 4 (PDF pp. 4) · Review: machine-drafted-source-checked

interpretive methods should be compared through partial and measurable performance rather than claims to perfect correctness

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 4, that interpretive methods should be compared through partial and measurable performance rather than claims to perfect correctness. The predictive frame accepts degrees of success and distinguishes consistency from closeness to intent. This is significant because a probabilistic task calls for evaluation metrics rather than all-or-nothing declarations of meaning. It connects to partial accuracy, measurement, probabilistic judgment, interpretive performance, empirical legal studies, model evaluation.

printed pp. 4 (PDF pp. 4) · Review: machine-drafted-source-checked

bias and variance provide a second way to evaluate interpretive models beyond the precision-accuracy distinction

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 4, that bias and variance provide a second way to evaluate interpretive models beyond the precision-accuracy distinction. Model complexity affects systematic error, instability, explainability, and predictability. This is significant because legal theories must account for how rules perform on new disputes, not only how well they fit one case. It connects to bias-variance tradeoff, model complexity, generalization, interpretive theory, explainability, predictability.

printed pp. 4 (PDF pp. 4) · Review: machine-drafted-source-checked

meaning prediction and purpose simulation are different interpretive modalities with different data and normative implications

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 4, that meaning prediction and purpose simulation are different interpretive modalities with different data and normative implications. Courts commonly blend them even though one models linguistic expression and the other models the parties themselves. This is significant because clarifying the target of prediction is necessary before choosing evidence or measuring accuracy. It connects to meaning, purpose, simulation, objective function, evidentiary data, normative theory.

printed pp. 4 (PDF pp. 4) · Review: machine-drafted-source-checked

modern doctrine generally states that interpretation seeks the parties' intention at the time of contract formation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 4–5, that modern doctrine generally states that interpretation seeks the parties' intention at the time of contract formation. Written language is then treated as the best evidence of that historical intention. This is significant because the consensus objective leaves unresolved how courts should infer intent from available sources. It connects to formation intent, objective interpretation, written agreement, historical inquiry, contract doctrine, evidence.

printed pp. 4-5 (PDF pp. 4-5) · Review: machine-drafted-source-checked

contract theory may posit a correct interpretive answer even though courts disagree about how to operationalize the search

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 5, that contract theory may posit a correct interpretive answer even though courts disagree about how to operationalize the search. The controversy concerns the model used to estimate the answer rather than the abstract aspiration to recover intent. This is significant because agreement on ends can conceal deep disagreement about inputs, costs, and error. It connects to correct answer, operationalization, interpretive methods, party intent, adjudication, contract theory.

printed pp. 5 (PDF pp. 5) · Review: machine-drafted-source-checked

textualism uses a relatively restricted evidentiary set centered on the final document

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 5, that textualism uses a relatively restricted evidentiary set centered on the final document. It relies on text, judicial semantic knowledge, dictionaries, and interpretive canons within the contract's four corners. This is significant because textualism is best understood as constrained contextual inquiry rather than context-free word worship. It connects to textualism, four corners, dictionaries, interpretive canons, semantic knowledge, evidence limits.

printed pp. 5 (PDF pp. 5) · Review: machine-drafted-source-checked

textualists consider internal context but ordinarily reach outside the document only for unavoidable ambiguity or voidability

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 5, that textualists consider internal context but ordinarily reach outside the document only for unavoidable ambiguity or voidability. Their disagreement with contextualists concerns the permissible range and timing of contextual evidence. This is significant because the debate is about institutional boundaries around context, not whether words are ever situated. It connects to internal context, ambiguity, parol evidence, textualism, voidability, evidentiary sequence.

printed pp. 5 (PDF pp. 5) · Review: machine-drafted-source-checked

contextualism treats the contractual text as important but admits a much broader body of potentially probative evidence

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 5, that contextualism treats the contractual text as important but admits a much broader body of potentially probative evidence. Special-community usage, trade usage, course of dealing, course of performance, and formation context can all inform the inference. This is significant because contextualism changes both the information available to courts and the complexity of adjudication. It connects to contextualism, trade usage, course of performance, course of dealing, formation context, parol evidence.

printed pp. 5 (PDF pp. 5) · Review: machine-drafted-source-checked

purposive interpretation asks why the parties adopted a term, while literal interpretation asks what the language means

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 6, that purposive interpretation asks why the parties adopted a term, while literal interpretation asks what the language means. For a fourteen-day delivery clause, a literalist parses when the clock starts, whereas a purposivist investigates the function of the deadline. This is significant because identical text can be analyzed through different target variables. It connects to purposivism, literalism, contract purpose, word meaning, delivery terms, interpretive target.

printed pp. 6 (PDF pp. 6) · Review: machine-drafted-source-checked

purposivism and literalism can combine with either textualism or contextualism

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 6, that purposivism and literalism can combine with either textualism or contextualism. A contextualist may use outside evidence to identify meaning or purpose, and a text can itself state its purpose. This is significant because evidentiary breadth and interpretive objective are separate dimensions of theory. It connects to purposivism, literalism, textualism, contextualism, two-dimensional taxonomy, contract evidence.

printed pp. 6 (PDF pp. 6) · Review: machine-drafted-source-checked

American scholars tend toward contextualism while courts tend toward textualism

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 6–7, that American scholars tend toward contextualism while courts tend toward textualism. The longstanding dispute also carries pragmatic, metaphysical, and sometimes political associations. This is significant because institutional practice may diverge from academic preference even when both use the language of party intent. It connects to courts, legal scholarship, formalism, contextualism, legal politics, institutional practice.

printed pp. 6-7 (PDF pp. 6-7) · Review: machine-drafted-source-checked

the linguistic model seeks word meaning, whereas the predictive model uses any reliable source to estimate what the parties wanted to mean

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 7, that the linguistic model seeks word meaning, whereas the predictive model uses any reliable source to estimate what the parties wanted to mean. Text, industry norms, and judicial knowledge all matter to the extent that they improve the estimate. This is significant because evidence should be selected by predictive contribution rather than category alone. It connects to linguistic meaning, predictive evidence, industry norms, judicial priors, party intent, model selection.

printed pp. 7 (PDF pp. 7) · Review: machine-drafted-source-checked

the predictive approach is bottom-up, empirical, and committed to fidelity rather than metaphysical purity

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 7, that the predictive approach is bottom-up, empirical, and committed to fidelity rather than metaphysical purity. It asks which process maximizes predictive accuracy after adjudication and error costs are included. This is significant because interpretive doctrine can be revised when evidence shows that another process performs better. It connects to bottom-up method, empirical jurisprudence, predictive fidelity, error costs, adjudication costs, adaptive doctrine.

printed pp. 7 (PDF pp. 7) · Review: machine-drafted-source-checked

the best interpretive method can depend on institutional strength

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 7, that the best interpretive method can depend on institutional strength. Weak institutions may need text-centered rules, while capable and trustworthy institutions can use broader evidence productively. This is significant because interpretive theory should be institution-sensitive rather than universal across courts. It connects to institutional competence, textualism, contextualism, judicial capacity, comparative institutions, legal design.

printed pp. 7 (PDF pp. 7) · Review: machine-drafted-source-checked

contract interpretation is probabilistic and uncertain even when doctrine speaks in categorical terms

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 7, that contract interpretation is probabilistic and uncertain even when doctrine speaks in categorical terms. The predictive frame openly treats judicial conclusions as estimates rather than metaphysical discoveries. This is significant because acknowledging uncertainty makes error analysis and model comparison possible. It connects to probability, uncertainty, judicial confidence, model comparison, party intent, legal epistemology.

printed pp. 7 (PDF pp. 7) · Review: machine-drafted-source-checked

contextualists often move from the true claim that context shapes meaning to the unsupported claim that broad extrinsic evidence must be admitted

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 7–8, that contextualists often move from the true claim that context shapes meaning to the unsupported claim that broad extrinsic evidence must be admitted. The motte-and-bailey diagnosis separates linguistic pragmatics from the design of an evidentiary regime. This is significant because doctrinal conclusions require comparative institutional evidence, not only observations about language. It connects to context dependence, extrinsic evidence, motte and bailey, contextualism, evidence law, institutional analysis.

printed pp. 7-8 (PDF pp. 7-8) · Review: machine-drafted-source-checked

every sentence operates within interpretive assumptions and a situation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 8, that every sentence operates within interpretive assumptions and a situation. Stanley Fish's point captures the uncontroversial linguistic insight that reference and meaning depend on context. This is significant because textualism should not be caricatured as belief in wholly context-free words. It connects to Stanley Fish, linguistic pragmatics, context, reference, textualism, interpretive assumptions.

printed pp. 8 (PDF pp. 8) · Review: machine-drafted-source-checked

language's indeterminacy does not logically entail admitting testimony, trade usage, or every circumstance surrounding formation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 8–9, that language's indeterminacy does not logically entail admitting testimony, trade usage, or every circumstance surrounding formation. Corbin's 'therefore' moves from a semantic proposition to a much stronger rule about judicial evidence. This is significant because a theory of meaning cannot substitute for a theory of adjudicative cost and reliability. It connects to Corbin, semantic indeterminacy, parol evidence, trade usage, legal entailment, evidentiary policy.

printed pp. 8-9 (PDF pp. 8-9) · Review: machine-drafted-source-checked

a jurisprudence requiring all context produces an infinite regress with no principled stopping point

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 9, that a jurisprudence requiring all context produces an infinite regress with no principled stopping point. If every circumstance requires its own surrounding circumstances, complete context is impossible. This is significant because real courts must use bounded evidence and a satisfactory rather than exhaustive account of meaning. It connects to infinite regress, bounded rationality, contextual evidence, judicial economy, linguistic theory, stopping rules.

printed pp. 9 (PDF pp. 9) · Review: machine-drafted-source-checked

even committed contextualists ultimately accept a good-enough probabilistic approximation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 9–10, that even committed contextualists ultimately accept a good-enough probabilistic approximation. Remote or implausible interpretations can be ignored despite theoretical relevance. This is significant because actual interpretive practice is empirical and pragmatic rather than an unlimited search for veritable meaning. It connects to good-enough standard, probabilistic approximation, pragmatism, contextualism, legal relevance, bounded inquiry.

printed pp. 9-10 (PDF pp. 9-10) · Review: machine-drafted-source-checked

the strongest case for contextualism is that context improves predictions of actual party intent

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 10, that the strongest case for contextualism is that context improves predictions of actual party intent. References to a computer or payment may incorporate the object or arrangement the parties discussed, not an abstract prototype. This is significant because contextualism needs no metaphysical premise when it can be defended through improved fidelity. It connects to contextualism, predictive accuracy, implicit context, UCC, party psychology, commercial meaning.

printed pp. 10 (PDF pp. 10) · Review: machine-drafted-source-checked

the strongest case for textualism is that a cheap, constrained method can be sufficiently accurate

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 10, that the strongest case for textualism is that a cheap, constrained method can be sufficiently accurate. Plain meaning may perform well enough under practical constraints to make broader inquiry not worth its cost. This is significant because textualism is an empirical claim about cost-adjusted performance, not a claim that words possess context-free essence. It connects to textualism, plain meaning, cost-effectiveness, satisficing, predictive accuracy, judicial economy.

printed pp. 10 (PDF pp. 10) · Review: machine-drafted-source-checked

cost- and fairness-weighted predictive accuracy supplies a common benchmark for interpretive methods

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 10, that cost- and fairness-weighted predictive accuracy supplies a common benchmark for interpretive methods. The framework clears conceptual confusion and permits comparison across models rather than allegiance to labels. This is significant because a shared metric can expose which disagreements are empirical and which are normative. It connects to common benchmark, fairness, error costs, method comparison, empirical testing, contract interpretation.

printed pp. 10 (PDF pp. 10) · Review: machine-drafted-source-checked

interpretive debates use accuracy ambiguously, allowing both sides to claim victory

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 11, that interpretive debates use accuracy ambiguously, allowing both sides to claim victory. Machine-learning practice distinguishes repeatability from closeness to the target. This is significant because terminological precision is needed before courts and scholars can compare performance. It connects to accuracy, precision, terminology, machine learning, interpretive debate, performance metrics.

printed pp. 11 (PDF pp. 11) · Review: machine-drafted-source-checked

Varney v. Ditmars turns the phrase fair share of profits into a test of interpretive performance

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 11, that Varney v. Ditmars turns the phrase fair share of profits into a test of interpretive performance. An architect stayed after a bonus promise, was fired, and received no profit share because the majority found the phrase too indefinite. This is significant because the case exposes how a doctrine seeking certainty can select an outcome plainly inconsistent with the bargain's direction. It connects to Varney v. Ditmars, indefiniteness, profit sharing, employment contract, fair share, contract remedies.

printed pp. 11 (PDF pp. 11) · Review: machine-drafted-source-checked

the Varney majority chose the only allocation certainly inconsistent with an agreement to share some profits

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 11–12, that the Varney majority chose the only allocation certainly inconsistent with an agreement to share some profits. Denying the claim gave the employer one hundred percent even though the parties plainly contemplated a positive employee share. This is significant because formal certainty can maximize error rather than minimize it. It connects to formalism, indefinite terms, zero recovery, party intent, accuracy, judicial error.

printed pp. 11-12 (PDF pp. 11-12) · Review: machine-drafted-source-checked

Varney can be reconstructed as a penalty default sacrificing current accuracy to induce clearer future drafting

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 12, that Varney can be reconstructed as a penalty default sacrificing current accuracy to induce clearer future drafting. On this account, nonenforcement warns later parties that vague bonus terms will fail. This is significant because a seemingly mistaken interpretation may pursue dynamic incentives rather than case-specific fidelity. It connects to penalty defaults, drafting incentives, dynamic effects, indefiniteness, future contracts, interpretive policy.

printed pp. 12 (PDF pp. 12) · Review: machine-drafted-source-checked

the proposed penalty default in Varney targets the wrong drafter when employers control employment agreements

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 12, that the proposed penalty default in Varney targets the wrong drafter when employers control employment agreements. An employer benefits from vagueness if indefiniteness eliminates its duty to share profits. This is significant because information-forcing rules must be assigned against the party able and motivated to clarify. It connects to information forcing, employer drafting, penalty default, strategic ambiguity, employment law, incentives.

printed pp. 12 (PDF pp. 12) · Review: machine-drafted-source-checked

a textualist court could enforce a rough general understanding of fair share even if it misses the industry's exact custom

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 12, that a textualist court could enforce a rough general understanding of fair share even if it misses the industry's exact custom. A semi-arbitrary twenty-eighty division would be inaccurate relative to a stipulated ten-ninety norm but directionally faithful to a sharing promise. This is significant because partial accuracy can dominate the perfectly certain zero-share result. It connects to textual priors, rough justice, partial accuracy, profit allocation, industry custom, contract enforcement.

printed pp. 12 (PDF pp. 12) · Review: machine-drafted-source-checked

contextual evidence may bring a court closer to industry custom while also widening the range of possible outcomes

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 12–13, that contextual evidence may bring a court closer to industry custom while also widening the range of possible outcomes. Testimony and practice can reveal the true split but can also mislead through motivated or unreliable proof. This is significant because more information may improve expected closeness while reducing predictability. It connects to contextual evidence, industry custom, motivated testimony, outcome distribution, accuracy, precision.

printed pp. 12-13 (PDF pp. 12-13) · Review: machine-drafted-source-checked

textualism can be precise even when biased away from the parties' likely intent

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 13, that textualism can be precise even when biased away from the parties' likely intent. A narrow method clusters decisions around a stable twenty-eighty allocation that parties can anticipate. This is significant because predictability is a distinct institutional benefit rather than proof of substantive correctness. It connects to textualism, precision, predictability, bias, precedent, planning.

printed pp. 13 (PDF pp. 13) · Review: machine-drafted-source-checked

contextualism can be more accurate on average but less precise across cases

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 13, that contextualism can be more accurate on average but less precise across cases. Some decisions reach the stipulated true custom while others scatter far from both truth and one another. This is significant because fidelity and consistency may move in opposite directions. It connects to contextualism, accuracy, variance, judicial discretion, outcome scatter, predictability.

printed pp. 13 (PDF pp. 13) · Review: machine-drafted-source-checked

precision alone can be maximized by an arbitrary rule

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 13–14, that precision alone can be maximized by an arbitrary rule. Always ruling for one side is perfectly repeatable but has no necessary connection to party intent. This is significant because certainty is valuable only when paired with at least reasonable accuracy. It connects to mechanical rules, certainty, precision, arbitrariness, rule of law, party intent.

printed pp. 13-14 (PDF pp. 13-14) · Review: machine-drafted-source-checked

interpretive methods should be compared along a frontier where both precision and accuracy can improve

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 14, that interpretive methods should be compared along a frontier where both precision and accuracy can improve. The choice is broader than fantastically certain textualism versus messy but accurate contextualism. This is significant because hybrid methods may dominate the corner solutions without requiring an all-or-nothing school. It connects to Pareto frontier, hybrid interpretation, textualism, contextualism, method design, performance improvement.

printed pp. 14 (PDF pp. 14) · Review: machine-drafted-source-checked

the desired balance between precision and accuracy depends on normative commitments

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 14, that the desired balance between precision and accuracy depends on normative commitments. Equal-treatment theories value clustered outcomes, while individualized-merit theories value closeness to each pair's actual intent. This is significant because interpretive design embeds a theory of justice as well as a prediction model. It connects to equal treatment, individual merit, legal justice, precision, accuracy, normative choice.

printed pp. 14 (PDF pp. 14) · Review: machine-drafted-source-checked

commercial parties may value precision because it supports planning and settlement

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 14, that commercial parties may value precision because it supports planning and settlement. Predictable interpretations reduce dispute probability and permit pricing, but only while outcomes remain reasonably faithful. This is significant because textualism's administrability benefit has a limit when systematic error blocks desired bargains. It connects to commercial planning, settlement, litigation costs, textualism, predictability, party autonomy.

printed pp. 14 (PDF pp. 14) · Review: machine-drafted-source-checked

risk neutrality does not make interpretive bias irrelevant

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 14–15, that risk neutrality does not make interpretive bias irrelevant. Schwartz and Scott emphasize being correct on average, but choosing textualism can move the mean outcome as well as alter its variance. This is significant because a stable decision rule can systematically misprice or misenforce contracts even for sophisticated actors. It connects to risk neutrality, Schwartz and Scott, mean error, textualism, sophisticated parties, contract design.

printed pp. 14-15 (PDF pp. 14-15) · Review: machine-drafted-source-checked

contextualism's greater evidentiary breadth may improve the mean accuracy of interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 15, that contextualism's greater evidentiary breadth may improve the mean accuracy of interpretation. If additional evidence reveals industry-specific intent, its advantage is not merely occasional correction within the same average. This is significant because conceding contextual accuracy weakens a defense of textualism based solely on unbiased averages. It connects to contextual accuracy, mean outcome, industry evidence, method comparison, interpretive bias, contract scholarship.

printed pp. 15 (PDF pp. 15) · Review: machine-drafted-source-checked

textualism retains a distinct advantage in predictability even if contextualism is more accurate

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 15, that textualism retains a distinct advantage in predictability even if contextualism is more accurate. More clustered decisions facilitate settlement and reduce disputes. This is significant because method choice involves real institutional tradeoffs rather than a declaration that one school owns accuracy. It connects to predictability, settlement, dispute reduction, textualism, contextualism, institutional tradeoff.

printed pp. 15 (PDF pp. 15) · Review: machine-drafted-source-checked

additional extrinsic evidence can have negative marginal informational value

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 15–16, that additional extrinsic evidence can have negative marginal informational value. Parol evidence may be forged, inaccurate, self-serving, confusing, or unduly costly. This is significant because more data does not automatically improve a legal prediction when input quality is poor. It connects to parol evidence, information quality, evidence law, contextualism, marginal value, judicial error.

printed pp. 15-16 (PDF pp. 15-16) · Review: machine-drafted-source-checked

evidence law already recognizes that relevant information can reduce decision quality

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 16, that evidence law already recognizes that relevant information can reduce decision quality. Rule 403 permits exclusion when probative value is outweighed by prejudice, confusion, delay, or cumulation. This is significant because legal doctrine itself embodies a model-selection principle that filters harmful inputs. It connects to Rule 403, probative value, prejudice, evidence filtering, model selection, judicial administration.

printed pp. 16 (PDF pp. 16) · Review: machine-drafted-source-checked

interpretive processes are models that transform evidentiary inputs into predicted outcomes

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 16, that interpretive processes are models that transform evidentiary inputs into predicted outcomes. Their complexity affects accuracy, explainability, and the ability of parties to anticipate decisions. This is significant because the number and weighting of features are institutional design choices. It connects to interpretive models, features, verdict prediction, model complexity, explainability, legal institutions.

printed pp. 16 (PDF pp. 16) · Review: machine-drafted-source-checked

textualism is relatively low complexity because it limits features to text and a small set of linguistic tools

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 16, that textualism is relatively low complexity because it limits features to text and a small set of linguistic tools. A narrower input space generally makes the method easier to understand and predict. This is significant because simplicity can be a source of institutional transparency as well as error. It connects to low-complexity models, textualism, feature selection, interpretive canons, explainability, predictability.

printed pp. 16 (PDF pp. 16) · Review: machine-drafted-source-checked

contextualism is high complexity because the parties can introduce an open-ended set of circumstances

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 16, that contextualism is high complexity because the parties can introduce an open-ended set of circumstances. Emails, conversations, photographs, trade practice, and other facts lack a fixed weighting procedure. This is significant because open feature sets can make outcomes responsive but difficult to reproduce or forecast. It connects to high-complexity models, contextualism, open-ended evidence, feature weighting, judicial discretion, predictability.

printed pp. 16 (PDF pp. 16) · Review: machine-drafted-source-checked

prediction error can be decomposed conceptually into bias, variance, and irreducible error

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 16–17, that prediction error can be decomposed conceptually into bias, variance, and irreducible error. The equation clarifies that reducing one component may increase another. This is significant because an interpretation method cannot be assessed from a single success or failure. It connects to mean squared error, bias, variance, irreducible error, model evaluation, contract doctrine.

printed pp. 16-17 (PDF pp. 16-17) · Review: machine-drafted-source-checked

a simple winter-umbrella rule illustrates underfitting and predictable bias

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 17, that a simple winter-umbrella rule illustrates underfitting and predictable bias. It is often wrong on sunny days but remains rational, transparent, and better than never carrying an umbrella. This is significant because a rule can be knowingly imperfect yet useful because its errors are stable and comprehensible. It connects to underfitting, simple rules, predictable error, bias, decision heuristics, explainability.

printed pp. 17 (PDF pp. 17) · Review: machine-drafted-source-checked

a highly detailed umbrella rule can fit historical weather perfectly and fail on new conditions

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 17, that a highly detailed umbrella rule can fit historical weather perfectly and fail on new conditions. Spurious features such as day of week or a weatherperson's shirt color create apparent accuracy without generalization. This is significant because case-specific detail can mistake noise for signal. It connects to overfitting, spurious correlation, variance, generalization, complex rules, historical data.

printed pp. 17 (PDF pp. 17) · Review: machine-drafted-source-checked

optimal model complexity lies between underfitting and overfitting

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 18, that optimal model complexity lies between underfitting and overfitting. As complexity rises, bias usually falls while variance rises, requiring a balance between predictable error and brittle specificity. This is significant because interpretive theory should search for a sweet spot rather than a pure school. It connects to model complexity, underfitting, overfitting, sweet spot, bias-variance tradeoff, hybrid rules.

printed pp. 18 (PDF pp. 18) · Review: machine-drafted-source-checked

interpretive decisions participate in a feedback loop among courts, precedents, and contract drafters

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 18, that interpretive decisions participate in a feedback loop among courts, precedents, and contract drafters. Courts consult prior meanings, and parties choose words in anticipation of future judicial interpretation. This is significant because a method's value includes how well its precedents guide future conduct, not only its resolution of one dispute. It connects to precedent, contract drafting, feedback loops, legal prediction, dynamic effects, interpretive rules.

printed pp. 18 (PDF pp. 18) · Review: machine-drafted-source-checked

radical textualism produces systematic bias by excluding information necessary to understand contractual language

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 18–19, that radical textualism produces systematic bias by excluding information necessary to understand contractual language. Every repeated misprediction under the narrow procedure adds to model bias. This is significant because simplification becomes counterproductive when it removes signal rather than noise. It connects to radical textualism, systematic bias, information exclusion, party meaning, model error, contract precedent.

printed pp. 18-19 (PDF pp. 18-19) · Review: machine-drafted-source-checked

contextual evidence can reduce textualist bias while introducing variance

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 18–19, that contextual evidence can reduce textualist bias while introducing variance. A more complex model captures relevant circumstances but makes outcomes more sensitive to facts and proof specific to prior cases. This is significant because reducing systematic error does not guarantee better generalization. It connects to contextualism, bias reduction, variance, extrinsic evidence, model complexity, generalization.

printed pp. 18-19 (PDF pp. 18-19) · Review: machine-drafted-source-checked

context-heavy precedents may overfit earlier disputes and mislead later courts

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 19, that context-heavy precedents may overfit earlier disputes and mislead later courts. A past definition of flood can be too bound to the circumstances in which it was produced. This is significant because precedent transfers models across cases, making generalization error a legal concern. It connects to overfit precedent, flood exclusions, insurance contracts, contextualism, generalization error, stare decisis.

printed pp. 19 (PDF pp. 19) · Review: machine-drafted-source-checked

requiring each contextualist court to start from a blank slate is unrealistic and duplicative

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 19, that requiring each contextualist court to start from a blank slate is unrealistic and duplicative. Humans use prior decisions, and parties need stable predictions when choosing words today. This is significant because a method that is accurate only for a Solomonic judge with unlimited evidence is not institutionally viable. It connects to tabula rasa, judicial capacity, precedent, duplication costs, contract planning, institutional realism.

printed pp. 19 (PDF pp. 19) · Review: machine-drafted-source-checked

intermediate interpretive approaches are preferable to radical textualism or extreme contextualism

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 19–20, that intermediate interpretive approaches are preferable to radical textualism or extreme contextualism. The goal is a method as simple as possible but not simpler, balancing bias and variance. This is significant because corner solutions sacrifice one kind of error to minimize another. It connects to intermediate rules, integrative interpretation, bias-variance balance, textualism, contextualism, method design.

printed pp. 19-20 (PDF pp. 19-20) · Review: machine-drafted-source-checked

bias and variance do not exhaust the values relevant to interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 19–20, that bias and variance do not exhaust the values relevant to interpretation. Certainty, litigation cost, drafting cost, fairness, and the direction of error also matter. This is significant because technical metrics must be nested within legal and normative evaluation. It connects to litigation costs, drafting costs, fairness, error direction, bias, variance.

printed pp. 19-20 (PDF pp. 19-20) · Review: machine-drafted-source-checked

meaning-focused textualism and contextualism share an interpretive paradigm even though they use different amounts of evidence

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 20, that meaning-focused textualism and contextualism share an interpretive paradigm even though they use different amounts of evidence. Both treat the evidentiary breadcrumbs as inputs for understanding the agreement. This is significant because a deeper alternative must change the target, not merely broaden the record. It connects to interpretive paradigm, textualism, contextualism, evidence, meaning, simulation.

printed pp. 20 (PDF pp. 20) · Review: machine-drafted-source-checked

Krell v. Henry asks whether a room-hire agreement was implicitly conditional on a coronation procession

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 20, that Krell v. Henry asks whether a room-hire agreement was implicitly conditional on a coronation procession. The renter paid a deposit for a balcony view, refused the balance after cancellation, and the court treated the procession as foundational. This is significant because the case exposes tension between literal obligation, implied condition, and allocation of unexpected loss. It connects to Krell v. Henry, frustration, implied condition, coronation cases, risk allocation, contract purpose.

printed pp. 20 (PDF pp. 20) · Review: machine-drafted-source-checked

the implied-condition reading of Krell is not compelled because the parties might have chosen a nonrefundable allocation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 20–21, that the implied-condition reading of Krell is not compelled because the parties might have chosen a nonrefundable allocation. Krell incurred opportunity and preparation costs, while Henry might still have used the apartment for a gathering. This is significant because shared awareness of purpose does not uniquely determine who bears cancellation risk. It connects to nonrefundable reservation, opportunity costs, frustration, risk allocation, party expectations, contract remedies.

printed pp. 20-21 (PDF pp. 20-21) · Review: machine-drafted-source-checked

the coronation court's cab-fare distinction illustrates the instability of implied-purpose analysis

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 21, that the coronation court's cab-fare distinction illustrates the instability of implied-purpose analysis. It said a cancelled sporting event would not excuse a known-purpose cab ride because the purpose was not the contract's foundation. This is significant because calling a purpose foundational can restate rather than solve the allocation problem. It connects to foundation of contract, cab fare, implied condition, purpose, doctrinal line drawing, frustration.

printed pp. 21 (PDF pp. 21) · Review: machine-drafted-source-checked

the embedded-option view treats termination rules as negotiated allocations of loss rather than all-or-nothing meaning

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 21, that the embedded-option view treats termination rules as negotiated allocations of loss rather than all-or-nothing meaning. Deposits, reservation fees, and subcompensatory liquidated damages reflect opportunity cost, risk aversion, information, and third-party commitments. This is significant because remedies can encode a nuanced fallback bargain that neither literalism nor implied-condition analysis captures. It connects to embedded options, liquidated damages, deposits, termination rights, risk aversion, contract remedies.

printed pp. 21 (PDF pp. 21) · Review: machine-drafted-source-checked

Krell may have covertly allocated the cancellation loss rather than simply declaring one side right

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 21, that Krell may have covertly allocated the cancellation loss rather than simply declaring one side right. By dismissing the deposit claim while excusing the unpaid balance, the court effectively divided the seventy-five-pound price two-thirds to one-third. This is significant because the remedial outcome can reveal a simulated compromise obscured by the court's interpretive language. It connects to loss sharing, remedial allocation, Krell v. Henry, deposit, judicial construction, embedded option.

printed pp. 21 (PDF pp. 21) · Review: machine-drafted-source-checked

what parties meant to say can differ from what they meant to do

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 21–22, that what parties meant to say can differ from what they meant to do. A textualist may see an absolute obligation and a contextualist an implied condition, while a purpose model asks which allocation best fits the parties' objectives and constraints. This is significant because linguistic meaning is an imperfect proxy for transactional purpose. It connects to meaning-purpose gap, textualism, contextualism, risk allocation, party objectives, simulation.

printed pp. 21-22 (PDF pp. 21-22) · Review: machine-drafted-source-checked

meaning interpretation and purpose interpretation are critically distinct even though commentators often blend them

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 22, that meaning interpretation and purpose interpretation are critically distinct even though commentators often blend them. Williston defines interpretation through word sense, Corbin through ideas induced by language, and purposive practice can still be described in semantic terms. This is significant because shared vocabulary can conceal different predictive targets. It connects to Williston, Corbin, word meaning, purposivism, interpretive definitions, conceptual conflation.

printed pp. 22 (PDF pp. 22) · Review: machine-drafted-source-checked

Bagchi's airport-ride hypothetical shows that relational context can change the authority to depart from literal promises

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 22–23, that Bagchi's airport-ride hypothetical shows that relational context can change the authority to depart from literal promises. A close friend may cancel after confidently learning that Bob prefers a new ride, while a paid driver may not infer beyond the agreement. This is significant because the reliability of purpose prediction can shape contractual obligation. It connects to contract and promise, intimates, strangers, airport ride, relational context, purpose inference.

printed pp. 22-23 (PDF pp. 22-23) · Review: machine-drafted-source-checked

intimates can act on evolving purposes because superior knowledge improves predictive accuracy

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 23, that intimates can act on evolving purposes because superior knowledge improves predictive accuracy. Alice can draw on Bob's history and preferences, while Charlie has only general experience and a brief interaction. This is significant because permission to override text may depend on the quality of the decisionmaker's model of the other person. It connects to predictive accuracy, relational knowledge, intimates, commercial strangers, contract obligations, information asymmetry.

printed pp. 23 (PDF pp. 23) · Review: machine-drafted-source-checked

purpose interpretation works by simulating how a party would respond to a counterfactual scenario

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 23, that purpose interpretation works by simulating how a party would respond to a counterfactual scenario. Alice builds a model of Bob from past interactions and probes it against the changed travel plan. This is significant because simulation identifies a distinct cognitive process rather than a looser form of semantic interpretation. It connects to counterfactual simulation, mental models, party preferences, past interactions, predictive reasoning, purpose.

printed pp. 23 (PDF pp. 23) · Review: machine-drafted-source-checked

the reliability of simulation can generate different normative duties

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 23–24, that the reliability of simulation can generate different normative duties. Alice's well-grounded model may justify unilateral adjustment, while Charlie remains bound to less person-dependent criteria. This is significant because epistemic competence and relational position can affect what conduct a promise permits. It connects to normative obligation, epistemic authority, relational contracts, textual fidelity, simulation, contract morality.

printed pp. 23-24 (PDF pp. 23-24) · Review: machine-drafted-source-checked

simulation is a humanistic exercise continuous with empathy and ordinary social understanding

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 24, that simulation is a humanistic exercise continuous with empathy and ordinary social understanding. People routinely model how others will react to gestures, apologies, persuasion, gifts, and plans. This is significant because technological vocabulary should not obscure the familiar interpersonal capacity underlying purpose inference. It connects to empathy, social cognition, mental simulation, humanism, interpersonal prediction, theory of mind.

printed pp. 24 (PDF pp. 24) · Review: machine-drafted-source-checked

theory of mind supplies an innate but unevenly distributed capacity for purpose simulation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 24, that theory of mind supplies an innate but unevenly distributed capacity for purpose simulation. The skill develops early, varies among people, and generally improves with knowledge of the person being modeled. This is significant because adjudicative reliability may depend on expertise and informational intimacy rather than doctrine alone. It connects to theory of mind, cognitive development, individual differences, predictive accuracy, adjudication, relational knowledge.

printed pp. 24 (PDF pp. 24) · Review: machine-drafted-source-checked

meaning search and purpose simulation are different models with different objective functions and datasets

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 24–25, that meaning search and purpose simulation are different models with different objective functions and datasets. Meaning search focuses on how people expressed goals, while simulation focuses on the goals they wanted to express. This is significant because choosing between them changes both the target and the relevant evidence. It connects to objective functions, datasets, meaning search, purpose simulation, evidence selection, model design.

printed pp. 24-25 (PDF pp. 24-25) · Review: machine-drafted-source-checked

purpose should generally receive priority over meaning when the two conflict in ordinary contract law

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 25, that purpose should generally receive priority over meaning when the two conflict in ordinary contract law. Words are downstream of transactional aims, and Krell illustrates a court willing to depart from surface meaning. This is significant because contract law is ordinarily concerned with enabling parties' projects rather than worshipping their formulations. It connects to purpose priority, party autonomy, contract goals, Krell v. Henry, purposive interpretation, legal meaning.

printed pp. 25 (PDF pp. 25) · Review: machine-drafted-source-checked

purpose simulation offers a more coherent way to identify and fill contractual gaps

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 25, that purpose simulation offers a more coherent way to identify and fill contractual gaps. Literal methods struggle to distinguish intentional silence from an unaddressed contingency and sometimes rely on metaphors such as a yawning void. This is significant because a model of the parties' projects can give concrete content to supplementation. It connects to gap filling, contractual silence, default rules, purpose, incomplete contracts, judicial supplementation.

printed pp. 25 (PDF pp. 25) · Review: machine-drafted-source-checked

purpose simulation faces serious objections concerning data, adjudicator burden, third-party reliance, precision, and accuracy

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 25–26, that purpose simulation faces serious objections concerning data, adjudicator burden, third-party reliance, precision, and accuracy. Judges receive conflicting motivated evidence and may differ widely in their reconstructions. This is significant because a normatively attractive target is not useful if institutions cannot estimate it reliably. It connects to information burden, judicial subjectivity, third-party reliance, motivated evidence, precision, accuracy.

printed pp. 25-26 (PDF pp. 25-26) · Review: machine-drafted-source-checked

successful cases show that simulation can sometimes be predictable, information-efficient, sensible, and accurate

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 26, that successful cases show that simulation can sometimes be predictable, information-efficient, sensible, and accurate. Krell, Bagchi's intimate, and commercial cases such as Nanakuli illustrate contexts in which purpose can be inferred without unlimited data. This is significant because institutional objections justify calibration rather than categorical rejection. It connects to Krell v. Henry, Nanakuli, commercial purpose, information efficiency, purpose inference, contextual adjudication.

printed pp. 26 (PDF pp. 26) · Review: machine-drafted-source-checked

commercial expertise should affect how heavily adjudicators rely on inferred purpose

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 26, that commercial expertise should affect how heavily adjudicators rely on inferred purpose. Posner's proposal gives experienced judges greater warrant to reason from business objectives. This is significant because method choice can be tailored to decisionmaker competence. It connects to commercial courts, judicial expertise, purpose inference, institutional competence, specialization, predictive accuracy.

printed pp. 26 (PDF pp. 26) · Review: machine-drafted-source-checked

purpose simulation requires a broader evidentiary dataset than language interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 26, that purpose simulation requires a broader evidentiary dataset than language interpretation. Negotiation choices, commercial setting, party characteristics, recitals, and declarations of purpose help reconstruct objectives, while definitions improve linguistic accuracy. This is significant because contract design can supply different data depending on which model a future court will use. It connects to contract recitals, definition sections, negotiation history, commercial context, purpose evidence, contract design.

printed pp. 26 (PDF pp. 26) · Review: machine-drafted-source-checked

human data-processing limits place an upper bound on purpose-based adjudication

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 26–27, that human data-processing limits place an upper bound on purpose-based adjudication. Industry arbitrators may handle thick records in specialized domains, but ordinary judges and juries cannot integrate unlimited information. This is significant because the feasibility of an interpretive ideal depends on adjudicative technology. It connects to bounded rationality, data processing, arbitration, industry expertise, judicial capacity, purpose simulation.

printed pp. 26-27 (PDF pp. 26-27) · Review: machine-drafted-source-checked

language models may weaken the data-processing objection to purposive interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 27, that language models may weaken the data-processing objection to purposive interpretation. Even a modest role in summarizing and organizing evidence changes how much contextual information adjudicators can realistically use. This is significant because new tools can shift the optimal interpretive method by changing institutional cost constraints. It connects to language models, legal technology, evidence summarization, judicial capacity, purpose simulation, interpretive change.

printed pp. 27 (PDF pp. 27) · Review: machine-drafted-source-checked

advances in AI modeling may move future contract law from linguistic meaning toward party purposes

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 27, that advances in AI modeling may move future contract law from linguistic meaning toward party purposes. If subjectivity and processing limits decline, purpose simulation becomes comparatively more feasible and attractive. This is significant because interpretive doctrine is endogenous to available technology. It connects to AI adjudication, purposive interpretation, institutional constraints, legal evolution, modeling parties, contract doctrine.

printed pp. 27 (PDF pp. 27) · Review: machine-drafted-source-checked

contract specificity is a function of interpretive technology and interpretive technology responds to drafting

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 27, that contract specificity is a function of interpretive technology and interpretive technology responds to drafting. Better purpose inference can reduce the need for exhaustive state-contingent language. This is significant because the cost of writing a good contract depends on what future adjudicators can reconstruct. It connects to contract specificity, interpretive technology, drafting costs, state contingencies, legal services, institutional feedback.

printed pp. 27 (PDF pp. 27) · Review: machine-drafted-source-checked

better purposive interpretation could lower barriers to complex transactions and scale relational contracting

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 27, that better purposive interpretation could lower barriers to complex transactions and scale relational contracting. Parties now spend heavily on deliberation, careful language, and lawyers to produce a good contract, even though nominal formation is cheap. This is significant because reliable reconstruction may expand who can coordinate through nuanced agreements. It connects to access to contracting, relational contracts, transaction costs, legal drafting, contract innovation, scale.

printed pp. 27 (PDF pp. 27) · Review: machine-drafted-source-checked

contract interpretation is prediction across time and should be evaluated like other predictive enterprises

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 28, that contract interpretation is prediction across time and should be evaluated like other predictive enterprises. The comparison invites systematic attention to inputs, targets, model complexity, and error. This is significant because legal interpretation need not remain insulated from general knowledge about prediction. It connects to temporal prediction, model evaluation, contract interpretation, machine learning, legal methodology, interdisciplinary research.

printed pp. 28 (PDF pp. 28) · Review: machine-drafted-source-checked

legal interpretation has been strikingly unempirical about the performance of its techniques

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 28, that legal interpretation has been strikingly unempirical about the performance of its techniques. Courts and scholars debate methods without routinely measuring how accurately they recover intent or how consistently they operate. This is significant because doctrinal confidence should be tested against evidence rather than sustained only by theory and intuition. It connects to empirical legal studies, interpretive validation, textualism, contextualism, legal methodology, measurement.

printed pp. 28 (PDF pp. 28) · Review: machine-drafted-source-checked

precision, accuracy, bias, and variance give legal scholars a vocabulary for asking answerable questions about interpretation

Professor Yonathan Arbel claims, in the article “Time and Contract Interpretation: Lessons from Machine Learning” on pages 28, that precision, accuracy, bias, and variance give legal scholars a vocabulary for asking answerable questions about interpretation. The chapter closes by asking how accurate textualism is, how precise contextualism is, and what bias each creates. This is significant because conceptual clarity is a precondition for credible empirical comparison and institutional reform. It connects to research agenda, precision, accuracy, bias, variance, contract interpretation.

printed pp. 28 (PDF pp. 28) · Review: machine-drafted-source-checked

Machine Files

Full Text Entry Point

The cleaned full text is exposed at fulltext_clean.txt, with fulltext_raw.txt preserved for audit. The compatibility path fulltext.txt points to the cleaned text. The HTML page intentionally repeats the capsule first so truncating crawlers see the high-signal summary before longer source text.