I should imagine some of you will be surprised to read these excerpts from an RAeS Upset and Loss of Control document.
A full stall is recoverable if resolved early in the stall. A prolonged stall condition can lead to an unrecoverable deep stall or a developed spin.
A simple swept and tapered wing will tend to stall first at the wingtips because the high loading outboard, due to taper and is aggravated by sweep-back. The boundary layer outflow also resulting from sweep reduces the lift capability near the tips and further worsens the situation. This causes a loss of lift outboard (and therefore aft) which produces pitch up. A lot of design sophistication is needed, including the use of camber and twist, leading edge breaker strips, fences etc., to suppress this inherent raw quality and cause an inboard section to stall first, so that the initial pitch tendency is a more desirable nose down. However, when a highly developed swept wing is taken beyond its initial stalling incidence, the tips may still become fully stalled before the inner wing in spite of the initial separation occurring inboard. The wing will then pitch up.