A legal custom is the established pattern of behavior within a particular social setting. A claim can be carried out in defense of "what has always been done and accepted by law".

Customary law (also, consuetudinary or unofficial law) exists where:

Most customary laws deal with standards of the community that have been long-established in a given locale. However, the term can also apply to areas of international law where certain standards have been nearly universal in their acceptance as correct bases of action – for example, laws against piracy or slavery (see hostis humani generis). In many, though not all instances, customary laws will have supportive court rulings and case law that have evolved over time to give additional weight to their rule as law and also to demonstrate the trajectory of evolution (if any) in the judicial interpretation of such law by relevant courts.

A central issue regarding the recognition of custom is determining the appropriate methodology to know what practices and norms actually constitute customary law. It is not immediately clear that classic Western theories of jurisprudence can be reconciled in any useful way with conceptual analyses of customary law, and thus some scholars (like John Comaroff and Simon Roberts) have characterized customary law norms in their own terms. Yet, there clearly remains some disagreement, which is seen in John Hund's critique of Comaroff and Roberts' theory, and preference for the contributions of H. L. A. Hart. Hund argues that Hart's The Concept of Law solves the conceptual problem with which scholars who have attempted to articulate how customary law principles may be identified, defined, and how they operate in regulating social behavior and resolving disputes.

Customary law refers to the set of customs, practices, and beliefs that a community accepts as binding rules of conduct.

Comaroff and Roberts' famous work, "Rules and Processes", attempted to detail the body of norms that constitute Tswana law in a way that was less legalistic (or rule-oriented) than had Isaac Schapera. They defined "mekgwa le melao ya Setswana" in terms of Casalis and Ellenberger's definition: melao therefore being rules pronounced by a chief and mekgwa as norms that become customary law through traditional usage. Importantly, however, they noted that the Tswana seldom attempt to classify the vast array of existing norms into categories and they thus termed this the 'undifferentiated nature of the normative repertoire'. Moreover, they observe the co-existence of overtly incompatible norms that may breed conflict, either due to circumstances in a particular situation or inherently due to their incongruous content. This lack of rule classification and failure to eradicate internal inconsistencies between potentially conflicting norms allows for much flexibility in dispute settlement and is also viewed as a 'strategic resource' for disputants who seek to advance their own success in a case. The latter incongruities (especially inconsistencies of norm content) are typically solved by elevating one of the norms (tacitly) from 'the literal to the symbolic. This allows for the accommodation of both as they now theoretically exist in different realms of reality. This is highly contextual, which further illustrates that norms cannot be viewed in isolation and are open to negotiation. Thus, although there are a small number of so-called non-negotiable norms, the vast majority are viewed and given substance contextually, which is seen as fundamental to the Tswana.

Comaroff and Roberts describe how outcomes of specific cases have the ability to change the normative repertoire, as the repertoire of norms is seen to be both in a state of formation and transformation at all times. These changes are justified on the grounds that they are merely giving recognition to de facto observations of transformation. Furthermore, the legitimacy of a chief is a direct determinant of the legitimacy of his decisions. In the formulation of legislative pronouncements, as opposed to decisions made in dispute resolution, the chief first speaks of the proposed norm with his advisors, then council of headmen, then the public assembly debate the proposed law and may accept or reject it. A chief can proclaim the law even if the public assembly rejects it, but this is not often done; and, if the chief proclaims the legislation against the will of the public assembly, the legislation will become melao, however, it is unlikely that it will be executed because its effectiveness depends on the chief's legitimacy and the norm's consistency with the practices (and changes in social relations) and will of the people under that chief.

Regarding the invocation of norms in disputes, Comaroff and Roberts used the term, "paradigm of argument", to refer to the linguistic and conceptual frame used by a disputant, whereby 'a coherent picture of relevant events and actions in terms of one or more implicit or explicit normative referents' is created. In their explanation, the complainant (who always speaks first) thus establishes a paradigm the defendant can either accept and therefore argue within that specific paradigm or reject and therefore introduce his or her own paradigm (usually, the facts are not contested here). If the defendant means to change the paradigm, they will refer to norms as such, where actually norms are not ordinarily explicitly referenced in Tswana dispute resolution as the audience would typically already know them and just the way one presents one's case and constructs the facts will establish one's paradigm. The headman or chief adjudicating may also do same: accept the normative basis implied by the parties (or one of them), and thus not refer to norms using explicit language but rather isolate a factual issue in the dispute and then make a decision on it without expressly referring to any norms, or impose a new or different paradigm onto the parties.