At Digital Web Magazine, Didier Hilhorst talks about the Apples and Oranges of user research and design.

User research--such as usability testing--is, without a doubt, imperative, but it certainly isn' design. It identifies problems, but doesn't, except maybe at the most detailed level, suggest adequate solutions. Designers have to visualize and refine broad and detailed solutions, while user researchers supposedly extract facts from identified situations. The two fields quarrel when, rather than representing or reporting facts, user researchers put forward solutions.

...

Designers have to deal with an assortment of constraints, such as what the technology will allow or the time frame in which they receive information. Similarly, user researchers may be aware of the role of design, but in effect fail to understand the constraints designers have to work with. Therefore the problem seems to be of a structural nature, rather than rooted exclusively in stubbornness or disrespect.

One way for me (neither a designer nor a user researcher) to relate to this is through short story writing. I write a story, polish it and make it shine (so I hope). I give it to a group of intelligent readers who will give me feedback. They can identify places in the story that they had problems (didn't convince me, hated this character, have no idea what's going on here) and they can even suggest solutions, but the 'real' solution to the issues they uncover must be mine. Like the designers, I know the constraints I'm dealing with and ultimately what I'm trying to do.

...there is sometimes an absence of, or limited attention to, design in the early stages of the process. Let me say this loud and clear: design is not optional. Design is not simply retro-fitting elements of style and aesthetics. More often than not design seems to be closer to reverse engineering than actually being an integral part of the process. Failing to acknowledge design in early stages is not taking your product, customers or business seriously.

Related to the quote above, one thing that's puzzled me for years is how songwriting teams work. I can't imagine, though I've tried, how the words and the music can be separated such that one person can write the words and one the tune. To me, they can't exist, one without the other. And yet they do. Creating songs and words is an intricate collaborative project which requires communication and trust. It's most successful when tune is not privileged over words and vice versa. In the same way, even though user research seems safely 'objective' and therefore easier to elevate to a position of greater importance, design can't be neglected or tacked on later. The result is stronger when it is integral to usability, content, and focus.