Category: E-Discovery and Legal Holds

Culling

Also known as: Weeding, Data culling, Data reduction

Simply put

In recordkeeping and e-discovery, culling is the selective removal of items from a larger collection so that only the relevant or worthwhile material remains. In records management it often means weeding out short-term, transitory, or non-record material from a set of files, while in electronic discovery it means reducing a large body of collected data down to the material that may matter for a legal matter. The everyday sense of the word, and its long-standing use in fields such as agriculture and photography, refers more generally to selecting and removing individuals or items from a group.

Formal definition

Culling denotes the deliberate reduction of a collection by removing items that fall outside defined criteria, though the criteria differ markedly by context and jurisdiction. In records management, culling (sometimes called weeding) typically refers to identifying and removing non-record, transitory, duplicate, or unscheduled material from a file or aggregation, and it should not be equated with authorized disposition or destruction of scheduled records, which requires an approved retention and disposition authority. In electronic discovery, culling refers to the defensible reduction of a collected data set by applying filters such as deduplication, de-NISTing (removal of known system files), date-range limits, file-type restrictions, and keyword or relevance screening, with the aim of narrowing the population to material that is potentially responsive, privileged, or otherwise within scope; in the United States, such reduction is often assessed against proportionality expectations of the sort reflected in the Federal Rules of Civil Procedure, though the governing standards vary by jurisdiction. Culling in these contexts is defined by relevance, responsiveness, privilege, record status, or scheduling criteria rather than by quality or desirability alone, and it is distinct from classification, retention, transfer, and formal disposition.

Why it matters

Culling sits at the point where volume meets obligation. In records management, organisations accumulate large quantities of transitory, duplicate, and non-record material alongside the records that document business activity. Weeding this material out helps staff locate authoritative records, reduces storage and management costs, and keeps aggregations coherent. However, culling must be exercised with care: removing something that is in fact a scheduled record, or destroying material that is subject to a legal hold, can undermine evidential value and create legal exposure. For this reason culling should be distinguished from authorised disposition, which in many jurisdictions requires an approved retention and disposition authority and cannot be performed on scheduled records simply because they seem unwanted.

Who it's relevant to

Records managers and information governance officers

Culling is a routine part of maintaining orderly files and aggregations, allowing non-record and transitory material to be weeded out while authoritative records are preserved and managed under their retention rules. These professionals must be careful to distinguish discretionary weeding from formal disposition, since scheduled records typically require an approved disposition authority before they can be destroyed or transferred.

E-discovery and litigation support teams

For those managing electronic discovery, culling is a core data-reduction stage that narrows a collected population using techniques such as deduplication, de-NISTing, date and file-type filtering, and keyword or relevance screening. The reduction must be defensible and, in jurisdictions such as the United States, is often weighed against proportionality expectations, so decisions and methods should be documented.

Legal and compliance leads

Culling intersects directly with legal holds and evidentiary obligations. Compliance and legal staff need to ensure that material subject to a hold is not culled, that reduction of collected data can be justified as defensible and proportionate, and that record status and privilege criteria, not mere desirability, drive what is retained or removed. Requirements vary by jurisdiction and sector.

Archivists

Archivists may encounter culling when appraising records for transfer, distinguishing material warranting continued or permanent preservation from short-term or non-record material. This work should be governed by appraisal and scheduling criteria rather than by desirability alone, and it remains distinct from the broader disposition decisions that determine whether records are retained, transferred, or destroyed.

Inside Culling

Culling (records management sense)

The selective removal or weeding of material that does not warrant retention, such as short-term or transitory items, non-record material, duplicate copies, or content not covered by an approved retention schedule. In many jurisdictions, national and state archival guidance (for example, U.S. National Archives guidance) describes culling as a step in preparing records for transfer, scheduling, or disposition, ensuring that only material meriting continued retention is carried forward.

Culling (e-discovery sense)

The defensible reduction of an electronically stored information (ESI) collection to a more focused, reviewable set. In this context culling is typically driven by relevance, responsiveness, and privilege criteria rather than by intrinsic quality or desirability, and is applied after preservation and collection to narrow data for review and production.

Selection criteria

The basis on which material is removed. In records management this often turns on record status (record versus non-record or transitory), duplication, and coverage by a retention schedule. In e-discovery, criteria commonly relate to relevance and responsiveness to the matter, privilege, and other litigation-defined parameters. Confusing these two rationales can lead to conceptual error, so the applicable framework should be identified explicitly.

Common e-discovery culling techniques

Techniques frequently used to reduce ESI volumes include de-NISTing (removing known system and program files), deduplication (eliminating identical copies), and filtering by date range, file type, custodian, or keyword. These operate on the collected data set to lower review burden while preserving defensibility.

Relationship to disposition and destruction

Culling is not synonymous with destruction. Material identified through culling may be routed to various outcomes depending on organizational policy and applicable schedules, including deletion of transitory or non-record items, separation of duplicates, or exclusion from a review set. Disposition of scheduled records remains governed by approved authorities and should not be conflated with informal weeding.

Proportionality and defensibility

Culling decisions are expected to be documented and defensible. In e-discovery contexts in many jurisdictions, proportionality considerations (for example, those reflected in FRCP 26(b)(1) in U.S. federal practice) inform how far a data set may be narrowed relative to the needs of the matter. In records management, defensibility rests on consistent application of policy and schedules and on retaining evidence of the decisions made.

Common questions

Answers to the questions practitioners most commonly ask about Culling.

Is it true that 'culling' has no meaning in records management or information governance?

No. Culling has an established meaning in recordkeeping practice, where it typically refers to the selective removal or 'weeding' of short-term, non-record, duplicate, or unscheduled material from a collection or system. National archival authorities and archival literature recognize this usage. The term should not be dismissed as informal; rather, its precise meaning depends on context, and it is distinct from formal disposition carried out under an authorized retention schedule.

Does culling simply mean removing items based on quality or how desirable they are?

Not in a recordkeeping or e-discovery context. Framing culling as removal based on 'quality' or 'desirability' misrepresents the criteria that typically govern the practice. In records management, culling is usually guided by whether material has continuing value, is a duplicate, is transitory, or falls outside a retention schedule. In e-discovery, culling reduces a data set according to relevance, responsiveness, and privilege, together with objective filters, rather than subjective judgments about desirability.

What techniques are commonly used to cull a data set during e-discovery?

Common techniques often include de-NISTing (removing known system and program files that are unlikely to be relevant), deduplication, date-range filtering, file-type filtering, and keyword or search-term culling. These methods are typically applied to reduce a collected data set to material more likely to be relevant before review. The specific techniques used depend on the matter, the tools available, and the agreements or orders governing the discovery process, and defensibility often depends on documenting the criteria applied.

How does the duty of proportionality relate to culling in litigation?

In many jurisdictions, proportionality principles, such as those reflected in the U.S. Federal Rules of Civil Procedure at Rule 26(b)(1), inform how far a data set can be reduced and what effort is reasonable relative to the needs of the case. Culling is often justified on proportionality grounds, since reducing the volume of material to be reviewed can limit burden and cost. Because proportionality standards and their application vary by jurisdiction and matter, parties should confirm the applicable rules and any court expectations before relying on culling decisions.

How can an organization make its culling process defensible?

Defensibility typically rests on documenting the criteria and methods used, applying them consistently, and being able to explain and reproduce the decisions later. This often includes recording filters and search terms applied, retaining audit trails, and ensuring that culling does not inadvertently remove material subject to a legal hold or a continuing retention obligation. Because expectations differ by jurisdiction and by matter, organizations generally align their approach with applicable rules, any governing agreements, and internal policy.

How should culling be distinguished from formal disposition under a retention schedule?

Culling generally refers to the selective removal of short-term, duplicate, transitory, or unscheduled material, whereas formal disposition is the authorized action, which may include destruction, transfer, or permanent preservation, applied to records once their retention period has elapsed under an approved schedule. Treating culling as a substitute for scheduled disposition can create risk, since records still within a retention period or under legal hold should not be culled. Where the two overlap, organizations typically clarify in policy which actions require formal authorization and how each is recorded.

Common misconceptions

Culling and destruction mean the same thing.

Culling is the act of identifying and separating material that need not be carried forward; it precedes and informs disposition. The resulting outcome may be deletion, but may also be exclusion from a review set or separation of duplicates. Destruction of scheduled records remains a distinct, authority-governed step, so the two terms should not be treated as interchangeable.

Culling is governed by the same criteria in every context.

The rationale differs by context. In records management, culling typically targets transitory, non-record, duplicate, or unscheduled material. In e-discovery, it is driven by relevance, responsiveness, and privilege. Framing e-discovery culling around 'quality or desirability' misrepresents the relevance- and privilege-based criteria that actually apply.

Culling is informal deletion that requires no accountability.

Culling decisions are generally expected to be defensible and documented. In litigation contexts, proportionality principles (such as those reflected in FRCP 26(b)(1) in U.S. federal practice) and the need to preserve material under legal hold constrain what may be culled, and records-management culling should align with approved retention schedules rather than ad hoc judgment.

Best practices

Clarify at the outset whether you are performing records-management culling (removal of transitory, non-record, duplicate, or unscheduled material) or e-discovery culling (defensible reduction of an ESI collection), because the governing criteria and objectives differ.

Apply relevance, responsiveness, and privilege criteria for e-discovery culling, and use established techniques such as de-NISTing, deduplication, and date, file-type, custodian, or keyword filtering to reduce volumes defensibly.

Confirm that no legal hold or preservation obligation applies before culling material, since obligations vary by jurisdiction and matter and can override routine reduction.

Align records-management culling with approved retention schedules and organizational policy, treating unscheduled material with caution rather than assuming it can be removed.

Document the criteria, methods, and outcomes of culling so decisions are defensible and can be explained if questioned, and observe proportionality considerations where litigation is involved.

Keep culling distinct from final disposition and destruction, routing separated material to the appropriate outcome and ensuring that destruction of scheduled records follows the required authorities.