Philosophical work on values in science is held back by widespread ambiguity about how values bear on scientific choices. Here, I disambiguate several ways in which a choice can be value-laden and show that this disambiguation has the potential to solve and dissolve philosophical problems about values in science. First, I characterize four ways in which values relate to choices: values can motivate, justify, cause, or be impacted by the choices we make. Next, I put my proposed taxonomy to work, using it to clarify one version of the argument from inductive risk. The claim that non-epistemic values must play a role in scientific choices that run inductive risk makes most sense as a claim about values being needed to justify such choices. The argument from inductive risk is not unique: many philosophical arguments about values in science can be more clearly understood and assessed by paying close attention to how values and choices are related.

关于价值如何影响科学选择的普遍模糊性阻碍了有关科学价值的哲学工作。在这里，我消除了选择可以承载价值的几种方式，并表明这种消除歧义具有解决和化解有关科学价值的哲学问题的潜力。首先，我描述了价值观与选择相关的四种方式：价值观可以激励、证明、导致或受到我们做出的选择的影响。接下来，我将我提出的分类法付诸实践，用它来澄清归纳风险论证的一个版本。非认知价值必须在具有归纳风险的科学选择中发挥作用的主张，作为需要证明价值的主张最有意义这样的选择。来自归纳风险的论点并不是唯一的：通过密切关注价值观和选择之间的关系，可以更清楚地理解和评估许多关于科学价值观的哲学论点。