By analogy of the argument in Broekhuis (2013:xiv), one can distinguish between:

• general Afrikaans;
• user varieties of Afrikaans; and
• usage varieties of Afrikaans.

General Afrikaans is that form of Afrikaans that all Afrikaans speakers/writers, to a greater or lesser extent, have in common. Some of the manners of speaking/writing might be more common in certain groups or communities than in others, but are nonetheless generally understandable and/or recognisable in the larger Afrikaans community. It is therefore the most widespread and unmarked form of Afrikaans.

For example, the /ɛ/ in the word ek I is often pronounced in the south as [ɛ], but in the north more like [æ], without necessarily being characteristic of a specific dialect, geolect or sociolect (but rather of broad regions). Both [ɛ] and [æ] are therefore considered part of general Afrikaans.

In the spoken language (or even in the written form) of some communities, adjectives are more often suffixed with an attributive -e than in other communities. For example, dit is 'n warm·e dag it is a warm·ATTR day it is a warm day would be common in these communities, while newspapers, for example, might perhaps rather write dit is 'n warm dag. Although the latter case might have been traditionally considered the unmarked form, we consider both warm·e and warm part of general Afrikaans, since it is so widespread and does not cause any intelligibility problems in the larger Afrikaans community.

By contrast, in the regional varieties of Afrikaans, one finds phenomena that are actually limited to only a certain (rather small) group of people in a particular region. For example, the pronunciation of the /ɛ/ as [æ] in words such as seks sex, tjek cheque, and Checkers (the name of a supermarket chain) are largely limited to white Afrikaans speakers in the Free State and surrounding regions. As such, it can be regarded as a phenomenon that is limited to a particular user group, and it is therefore not a central part of general Afrikaans, but rather of the regional variety North-Eastern Afrikaans (i.e. user variety; see this list of user varieties).

Similarly, a word like nademaal forasmuch as, and an expression like ten laste lê to place a burden upon are strongly associated with language usage in the legal fraternity. They are not widespread, might be difficult to understand/process in the broader Afrikaans community, and are therefore typical of Regsafrikaans ( Legal Afrikaans), a specific usage variety (see this list of usage varieties).

In Taalportaal, general Afrikaans is mostly described, while descriptions of user varieties and usage varieties are limited to comments, references to literature where specific phenomena have been investigated, or the identification of topics for future research. Where existing literature on these user or usage varieties is available, it is usually mentioned in the text under Extra information, or in notes.

It is here premised that:

• at any given moment in time, the socio-political context determines the demographics of the work force in the mainstream media industry; and hence
• these demographics determine the characteristics of the language variety that is being used generally across various media; and hence
• this language variety is the variety that is being read and heard most commonly in the language community at that given moment in time; and hence
• this language variety could be considered the so-called "general variety" of the language; and
• a synchronic linguistic description of this variety would be the closest possible approximation to a generalising description of the language.

This general presupposition has the advantage that the descriptive scope of the grammar of a language is set simultaneously broad enough (e.g. not restricted to only the standardised variety of a language, or to prescriptive norms of the language), and narrow enough (e.g. to be still practically executable within limited physical resource parameters, to be surveyable by readers, and to be falsifiable). It also makes it possible to operationalise the endeavour to write a scientific grammar of a language, because you only need to specify (as is done later in this topic):

• the period that you are covering in your description;
• what you consider to be mainstream media; and
• what you consider to be data sources of these media.

This approach also provides the linguist with an extenuation (if not a pretext!) to focus on matters at hand – i.e. to produce a synchronic description of the grammatical constructions of a language – without the need to constantly and consistently account for socio-political factors that played a role in the development of the grammar, sociolinguistic variables (specifically race, gender, and age) that might influence grammaticality judgements of a particular construction, and socio-geographic variation in the grammar. Of course, from a usage-based perspective all of these factors and variables are undeniably central to both our broad understanding of language, and our detailed and in-depth explanation of a peculiar or unique construction. However, when writing a comprehensive, scientific grammar of a language, it is also important to "keep it real", i.e. to have a broad coverage of as many constructions as possible that the vast majority of mother-tongue speakers of the language will agree are part of the language they use every day.

To be clear, when there is referred to general Afrikaans on Taalportaal, it should not be confused with Standard Afrikaans. In Taalportaal, Standard Afrikaans is regarded as a usage variety, just like, for example, Pulpit Afrikaans, Legal Afrikaans, Radio Afrikaans, or Pavement Afrikaans. Our view is, therefore, that Standard Afrikaans is a variety used by certain users in certain contexts and registers. It is therefore not some idealized meta-form of Afrikaans, but rather a form of mostly written Afrikaans that is often used, for example, in edited newspaper texts, academic articles and annual reports. As such, Standard Afrikaans is part of general Afrikaans described in this project, but is in no way the only form of general Afrikaans.

Given the above presuppositions, the current version of the Afrikaans grammar in Taalportaal covers the time period from circa 1990 to 2020. Related to a historical-political time period, one could say that general Afrikaans is in the current version of Taalportaal equated to what can be called "post-Apartheid Afrikaans" (Van der Waal 2012), as it is being used in mainstream media.

Here mainstream media is defined simply as established outlets for broadcasting or (e)publishing. This definition has some implications:

• Using established outlets as a definitional criterion implies that we concentrate to a large degree on professional language usage – at least as reference point, or point of departure. In its strictest form it implies that most of the data that is being used, is copy-edited sources, e.g. printed magazine articles, or scripted news bulletins. Less strictly seen, it would imply that the data is produced by professional or semi-professional language users using language on somewhat recognised (established) platforms (outlets), e.g. internet blogs, or unscripted broadcasts. However, excluded as a primary source, is data produced by the general public on Facebook, Twitter, or in private correspondence.
• By including broadcasting and (e)publishing in the definition, both spoken and written language are covered. This would also include new media, like podcasting, vlogging, blogging, etc.
• Lastly, mainstream simply means established, or somewhat widely recognised media source, including smaller, independent, alternative media houses, websites, etc. We do not use mainstream media here in the sense of news conglomerates.

In practical terms, compare the following examples of typical data sources:

• Prototypical sources of general Afrikaans (included as primary data):
  • News articles, columns, and advertisements published in national, regional, or local newspapers and/or e-platforms, such as Netwerk24, Maroela Media and Vrye Weekblad
  • Articles published in magazines, and copy-edited newsletters or year-end reports, such as Taalgenoot and FAK Nuusbrief
  • Articles/blogs published on generally recognised websites, such as LitNet and VivA
  • Audio and text material from traditional or internet radio stations, like Radio Sonder Grense, Bok Radio, or Lekker FM
  • Audio-visual and text material from television broadcasters and websites, like kykNET and Afrikaans.com
  • Novels, drama texts, and non-fiction published by established publishers, like Protea Boekhuis, LAPA, and NB Publishers
  • Etc.
• Marginal sources of general Afrikaans (excluded as primary data, but included as secondary material):
  • Poetry published by established publishers
  • Novels, drama texts, and non-fiction that are self-published
  • Niche blogs on less widely-known websites, like WatKykJy and Watertandresepte vir Oud en Jonk
  • Small, independent podcasts, or individuals' YouTube channels
  • Etc.
• Excluded from general Afrikaans (and therefore in principle excluded from data, but perhaps only included as tertiary data for contrastive purposes):
  • Posts and comments from micro-blogging sites, like Facebook, Twitter, and Instagram
  • Posts and comments from micro-vlogging sites, like YouTube
  • Comments by end-users on news websites
  • Scholarly articles, theses, and dissertations
  • Learner data, such as homework or essays produced by either mother-tongue or non-mother-tongue learners
  • Music and lyrics
  • Etc.

Based on the above categories, we consider the following to be sources of examples and data, ordered in sequence of importance:

1. Electronic corpora with data types from different usage situations (where such data and corpora are available)
2. Electronic and other dictionaries, terminology lists, word lists, etc.
3. Other texts or sources to which the authors have access (e.g. printed material)
4. Existing linguistic literature on general Afrikaans, with examples contained therein
5. Own intuitions, observations and insights from the authors and other linguists
   • In an attempt to subscribe to an underlying usage-based philosophy, we try to minimise these latter kinds of data. However, we recognise that mother-tongue linguists of Afrikaans are potentially rich sources of information, as long as their own intuitions are subjected to rigorous peer-review and scrutiny.