Good call. The speculation isn't unfounded though, although it isn't appreciably strong here. Word choice comes substantially faster when the most efficient ex

Good call. The speculation isn't unfounded though, although it isn't appreciably strong here. Word choice comes substantially faster when the most efficient expression can be chosen without any self-monitoring. It's quite plausible that in this case you'll find less instances of tip of the tongue. While the experiment here relied on rare words, and I'm just one data point, but I get ToT probably once for every 2000 words I say. When rapid code-switching is consciously available, I can quickly avert it because I have fast access to good candidate words, which reduces the apparent occurrence. If ToT is taken as a speech disfluency from unresolved competition for word access, you could mitigate it simply by creating an imbalance between the signal strength such that winner takes all (in which case the subway example does demonstrate something useful, although as you point out, the variables are jumbled up). Somehow I got a deja vu just now. This has been discussed here before, or somewhere, or what I am writing is coming from something else I have read... I also realize that ToT cannot be reduced to simple competition, but it is also plausible that slow access and competition manifest similarly, i.e. obstructed speech flow, while partial semantic access. WEAVER++ tried to explain this and I liked it very much (obviously I don't know what's state of the art now).