Access is restricted
You are not permitted to access this course's announcements and schedule right now.
Explore comprehensive course materials and learning objectives designed for your growth.

Speaking is the section students improve least, because it is the only one you cannot practise silently. Reading and listening can be done alone on a train. Speaking requires you to talk out loud, badly, in front of someone who will tell you what went wrong — and almost nobody arranges that for themselves.
This lab arranges it. Fourteen people, one hour a week, everyone speaks every session, and every recording is scored against the official rubric.
There is no start and no finish. Join in any week; the material cycles through the four task types every fortnight and repeats indefinitely with new prompts. Students typically stay two to four quarters — long enough to move a band or two — and then stop. That is the intended shape and there is no penalty for leaving.
Billing is quarterly rather than monthly because meaningful speaking gains take about twelve weeks and a monthly cycle encourages people to quit in week five, which is exactly when it feels worst and is about to get better.
Ten minutes: warm-up in pairs, cameras on, no scoring, purely to get the language moving
Twenty minutes: two timed tasks, everyone recording simultaneously
Twenty minutes: three recordings played to the group and scored aloud against the rubric, by the group
Ten minutes: one targeted pronunciation or delivery point, chosen from what the recordings showed
Having your recording played and discussed is the feature people most want to skip and the one that moves scores fastest. Not because criticism helps — because scoring other people's answers against the rubric teaches you what the rubric means far quicker than being scored does.
By your fourth session you will be predicting your own band before I say it, and being right. That is the point at which independent practice starts working, and it is why students who stay a full quarter improve and students who attend twice do not.
Every session has an asynchronous equivalent: the same prompts, recorded on your own time, returned with audio feedback from me within two days. Around a third of the lab uses this permanently, usually because of time zones, and their gains are comparable.
Come to one session before subscribing. Speak in it — listening to other people be scored while saying nothing is not a preview of this lab, it is a preview of the wrong version of it.