<p>The proliferation of digital data and the increasing number of scholars using digital humanities methodologies have led to calls for tool developers to focus on interoperability and usability. Historically, a majority of tools that have been developed for digital humanities have been designed to meet the needs of one specific project or type of project, even if the general capabilities of the tool (e.g. n-gram analysis) could be applied more broadly. This is in part due to time and funding constraints, but workshop participants perceived that even a greater reliance on standards would improve the situation.</p> <p>Scholars felt that openness was critical for digital humanities tools-- if there are assumptions or filters built into a tool, scholars must know about it and factor that into their interpretation of the results. This need to assess the tool and its inner workings is arguably at odds with the emphatic desire for simpler, more intuitive user interfaces. Some scholars felt that the amount of time certain tools take to learn is not worth the payoff, particularly if it requires diverging from their pre-existing workflow. Previous negative experiences with complicated user interfaces were cited as a reason some scholars are reluctant about digital humanities-- these interfaces can lead a scholar to "fear ... how they might be changed, fear ... dependence, fear ... not knowing something" (Ex 3, 1c-D). However, other scholars noted with dismay that they tend to use all the features of a given software release, and are frequently frustrated when some of the peripheral features change or are eliminated in subsequent releases.</p> <p>The information from the workshops about what scholars want from tools is conflicting, and not particularly compatible with a monolithic set of "best practices", beyond using standards whenever possible to facilitate interoperability. While the needs and preferences of the end-user should play a guiding role when designing a tool, how this should be approached will vary significantly depending on the group of scholars the tool is intended for.</p>