Haskell was written as academia's response to Miranda's licensing fees. Were Miranda freeware (but not source available or Free Software), it probably would still be well-known. It's still taught in many universities.
I certainly agree that most languages should be libre, but clearly it's not necessary for success. k has brought billions of dollars of profit, but is the most proprietary of all languages. C# developers are the most common in the world, for some reason. VBA is popular. Excel has the most programmers. Mathematica makes millions. Matlab makes millions.
Miranda's cost was a big problem, but not the only one. The other core issue that lead to the creation of Haskell was Miranda's license that essentially prohibited using it as a tool for programming language research. For good reason, Turner (Miranda's creator) wanted to avoid the fragmentation of Miranda into different dialects. From [1]:
> [T]he easiest way to move forward was to begin with an existing language, and evolve it in whatever direction suited us. Of all the lazy languages under development, David Turner’s Miranda was by far the most mature. It was pure, well designed, fulfilled many of our goals, had a robust implementation as a product of Turner’s company, Research Software Ltd, and was running at 120 sites. Turner was not present at the meeting, so we concluded that the first action item of the committee would be to ask Turner if he would allow us to adopt Miranda as the starting point for our new language. After a brief and cordial interchange, Turner declined. His goals were different from ours. We wanted a language that could be used, among other purposes, for research into language features; in particular, we sought the freedom for anyone to extend or modify the language, and to build and distribute an implementation. Turner, by contrast, was strongly committed to maintaining a single language standard, with complete portability of programs within the Miranda community. He did not want there to be multiple dialects of Miranda in circulation and asked that we make our new language sufficiently distinct from Miranda that the two would not be confused. Turner also declined an invitation to join the new design committee [...] Haskell owes a considerable debt to Miranda, both for general inspiration and specific language elements that we freely adopted where they fitted into our emerging design.
over a long enough time frame, it seems his view is standard. nowadays, there's not really a haskell standard other than ghc. there's other haskells than ghc (e.g. ghcjs) but they're all forks of ghc. research is achieved by enabling extra features
(i guess eta isn't tracking ghc, but i think that's because of unviability rather than a specific intention to fork the language.)
I think the difference is that today nobody is preventing you from forcing GHC or adding features, while back in the early 1990s, Research Software Ltd, the company creating Miranda, would probably have prevented others from forking Miranda. 30 years ago, the value of open source and the network effect for programming languages was not widely understood. In particular, giving away from free and without restrictions a company's core IP was inconceivable for traditional businesses!
I certainly agree that most languages should be libre, but clearly it's not necessary for success. k has brought billions of dollars of profit, but is the most proprietary of all languages. C# developers are the most common in the world, for some reason. VBA is popular. Excel has the most programmers. Mathematica makes millions. Matlab makes millions.