Requirements, controls and evidence: what one row of a mapping claims
A mapping row claims a control covers a requirement; only evidence shows it works.

A mapping is read one row at a time, and each row joins three unlike things: a requirement — a framework outcome or standard requirement, a statement of what protection must be achieved — the control that satisfies it in a real system, and the evidence that shows the control works. The row is a claim made across two vocabularies never designed to meet: frameworks describe outcomes and withhold method, while a control exists only after somebody selected, tailored and implemented it.
Two consequences follow. The row is only as good as its traceability — an assessment of the control must lead back to the requirement that selected it, or the mapping is an opinion — and evidence is chosen with the control, because a claim whose evidence was never defined cannot be settled. A mapping table records coverage as a claim; it does not prove anything.
A requirement is not a control
The framework end states outcomes and withholds method: the NIST Cybersecurity Framework (CSF) 2.0 says of its Core that “These outcomes are not a checklist of actions to perform; specific actions taken to achieve an outcome will vary by organization and use case” (lines 243–245). The system end is the opposite: a system needs “a tailored set of controls and specific implementation details determined by applying the existing RMF process” (SP 800-37 Rev. 2, lines 535–536). A requirement says what protection is wanted; a control is the implemented thing that provides it, and the row asserts the second is sufficient for the first — a judgement, not a synonym.
Selecting and tailoring the control
Selection is not copying a baseline. SP 800-37 Rev. 2 describes the organisation’s own: tailored baselines and overlays “complement the NIST control baselines by providing an opportunity to add or eliminate controls to accommodate organizational requirements while continuing to protect information commensurate with risk” (lines 2379–2381). Tailoring fixes the control end and is where ownership splits. A control “that can be inherited by one or more information systems” is a common control (line 2449; SP 800-53 Rev. 5: “inheritable by multiple systems or programs”, lines 1899–1900), and where “a requirement may not be fully met by a common control” the remainder becomes a hybrid control provided at system level (lines 2460–2463). For inherited controls, “System owners are not required to assess common controls that are inherited by their systems … the responsibility of the common control providers” (lines 2481–2483). The row must therefore record where its evidence lives, not only that a control exists.
Making the row verifiable
the claim (declared): requirement → control → evidence
the proof (assessed): assessment result → traced back to the requirement
and not one-to-one:
one control can answer several requirements
one requirement can need several controls
A mapping becomes checkable only through assessment, and SP 800-53A Rev. 5 supplies the link: an assessment procedure is “a set of assessment objectives, each with an associated set of potential assessment methods and assessment objects” (lines 1027–1028), and the “determination statements are linked to the content of the control … to ensure the traceability of assessment results back to the fundamental control requirements” (lines 1030–1032). That clause is what lets the mapping be tested rather than asserted: the assessment result is traceable back to the requirement. Two attributes bound the test — “The depth and coverage of security and privacy evidence can affect the level of assurance in the functionality implemented” (lines 994–995), coverage being “the scope and breadth of assessment objects included in the assessments” (lines 1020–1022). Which methods and objects those are belongs to the L2 sheet on control objectives and verification criteria; this row uses only the traceability clause and the determination it produces.
Evidence is chosen with the control
SP 800-53 Rev. 5 places evidence at selection: “During control selection and implementation, it is important for organizations to consider the evidence (e.g., artifacts, documentation) that will be needed to support current and future control assessments” (lines 2063, 2065–2067). The row must eventually be settled: CA-2 requires controls to be assessed “to determine the extent to which the controls are implemented correctly, operating as intended, and producing the desired outcome” (lines 6355–6358). Choose the control without choosing what will show it works and the row states an intention. Collecting that evidence across a programme, reusing it, and judging the quality and freshness of one piece are separate concerns, owned elsewhere in this area.
The framework side of the same row
Frameworks publish the middle of the row themselves. CSF 2.0’s Informative References are “mappings that indicate relationships between the Core and various standards, guidelines, regulations, and other content” (lines 468–470), pointing “to sources of guidance on each outcome” (lines 172–173). They are partial and directional: “Some Informative References are narrower in scope than a Subcategory. For example, a particular control from SP 800-53 … may be one of many references needed to achieve the outcome described in one Subcategory” (lines 472–474). One outcome can need several controls — which is why a row must record its direction.
What a row cannot claim
A mapping is a claim of coverage, and coverage is not proof: a row says a control is believed to satisfy a requirement, while only assessment produces a finding, and SP 800-53A’s traceability clause is what attaches that finding back to the requirement. Two limits follow. The relation is not one-to-one in either direction — one control can answer several requirements, one requirement can need several controls — so a row that does not say which requirement it claims against, and in which direction, is ambiguous. And mapping does not create conformity: naming a control against a requirement is a statement of intent, not a result. The gap and its treatment plan, the record of included controls, and the evidence programme are separate concerns, and filling in the row does not discharge them.
Level and prerequisites
L4 — the assurance layer of governance: not the definition of a control objective or verification criterion (L2), not keeping one chain traceable (L3), but recording, across a framework, what covers what and what would show it. Prerequisites: the L1 vocabulary of controls and evidence, the L2 records, and the L3 chain from risk to control to owner to evidence.
Where to go next
- Knowledge — the Knowledge register this sheet belongs to.
- Cybersecurity Governance — the macro-area.
- Assurance — the node this sheet sits in.
References
- NIST — SP 800-53A Rev. 5, Assessing Security and Privacy Controls — the §2.4 assessment objectives, the traceability clause, the depth and coverage attributes, and the §3.3 determination wording.
- NIST — SP 800-53 Rev. 5, Security and Privacy Controls — the evidence-at-selection sentence, CA-2, and the common/system-specific control distinction.
- NIST — SP 800-37 Rev. 2, Risk Management Framework — organizationally-tailored baselines, overlays, and common controls and inheritance.
- NIST — The NIST Cybersecurity Framework (CSF) 2.0 — the Core’s outcomes and the Informative References that map them.