We can perhaps summarize our questions about relationships as follows:

(1) What variables are related to the extent and rate of language spread?

(2) With what variables are differences in the shapes of diffusion curves associated?

(3) What variables distinguish early adopters, late adopters, and nonadopters?

(4) To what extent do the answers to questions i‑3 above differ for the diffusion of innovations in general and from the diffusion of linguistic (e.g., phonological, syntactic, lexical) innovations in particular?

For each of these questions we can ask about the individual contribution of each variable as well as about the joint or cumulative contribution of all variables to the prediction of the criterion variable. Each of these questions can of course be asked for communication networks of varying sizes and types. We can collapse these four questions into a single summarizing question that serves as a framework for the study of language spread: who adopts what, when, why, and how?

This framework provides us with rubrics for topics of research that are relevant to language spread. Some of these topics are listed below.

1. Who

a. What characteristics (e.g., position within the communication network, need‑achievement, openness to change) distinguish adopters from nonadopters and early adopters from late adopters?

b. What characteristics (e.g., linguistic heterogeneity, size, complexity) distinguish communication networks through which a language spreads rapidly from networks through which it spreads slowly or not at all?

2. Adopts

a. Do different levels of adoption (awareness, evaluation, knowledge, usage) spread at different rates and to different extents?

b. Do the levels of adoption necessarily form a scale whereby awareness serves as a prerequisite for evaluation, evaluation for knowledge, and knowledge for usage?

3. What

a. What structural characteristics (e.g., diversity of the variety that must be learned, similarity to varieties in the potential adopter's verbal repertoire) are associated with differences in the extent and rate of adoption?

b. What functional characteristics (e.g., skills required, betweengroup versus within‑group interaction, interaction with respect to horizontal versus vertical integration) are associated with differences in the extent and rate of adoption?

4. When

a. How much time is required for a given language variety to be adopted by a given communication network under given conditions?

b. What characteristics are associated with differences in the shape of diffusion curves?

5. Where

a. What kinds of social interaction within what types of societal domain promote or retard the acceptance of innovation?

b. What are the characteristics of those individuals who are likely to be influential as agents of change with respect to what types of adopter and for what levels of adoption?

6. Why

a. What national and personal incentives for planners to promote (or hinder) language spread are related to differences in the extent and rate of spread?

b. What incentives for potential adopters to accept (or reject) a language variety are related to differences in the extent and rate of spread?

7. How

a. What language‑planning activities are most likely to be successful for different types of adopter at different levels of adoption, through different types of change agent, and under different sociopolitical and economic conditions?

While it may be difficult to operationalize some of the variables which are employed in these questions, particularly when studied in connection with large social aggregates, it seems worth the effort, inasmuch as such operationalization will enable us to compare the results of different studies of language spread to each other and to studies of diffusion more generally. We cannot know what variables will be revealing unless we develop a common framework for their use. The framework presented here is offered as a contribution toward that end.