Teacher quality "trumps almost everything else", says Professor Dylan Wiliam of the Institute of Education in London. Few would dispute that assertion: what is more contentious is his argument that, by training large numbers of new teachers in a short time, the Scottish Government will dilute the quality of the teaching force.
Anecdotal evidence suggests that the standard of new teachers has never been higher. But if these new teachers do not find immediate employment, they risk becoming demotivated and losing the skills they acquired in training.
Professor Wiliam suggests that, so important is it to put an excellent teacher in front of a class, it will raise pupil attainment more if a very good teacher teaches a class of 40 than an average or below-average teacher is in charge of a class half that size. His formula for raising pupil attainment significantly is also dependent upon teachers using formative assessment, not surprisingly since it is his "baby". While he accepts that Scotland has gone further down this road than other parts of the UK, he does not believe we have truly embedded this approach in teaching practice.
Here, perhaps, is where his class size argument springs a few holes. For, as Ronnie Smith, general secretary of the Educational Institute of Scotland, points out, formative assessment requires teachers to monitor individual pupils' progress - and it is surely easier for even an excellent teacher to track the progress of 20 pupils than of 40.
Not surprisingly, the EIS likes Professor Wiliam's idea of cutting class contact time to 15 hours per week. But such an aspiration would require more teachers, which brings us up against the problem of diluting teacher quality again. So yes, teacher quality does indeed "trump almost everything else".