I am with Femme on this. Her teacher made a mistake.
My high school teacher (John B Davis) taught us boys " When they give you voltage between two leads, as in " Vbe" , your voltmeter red lead (positive) goes to the first letter and the black one (negative) to second letter".
I note that Fairchild datasheet for 2N3906, a common pnp, gives Vbe in operating range as ~ -0.65 volts, note minus sign.
That means a meter connected as above would find base negative wrt emitter, which is forward biased(emitter positive wrt base).
VBE(sat) Base-Emitter Saturation Voltage IC = -10mA, IB = -1.0mA
IC = -50mA, IB = -5.0mA -0.65(min) -0.85(typ) -0.95(max) V
Now here's a confusion factor - Same datasheet gives emitter to base reverse breakdown voltage as also negative,
V(BR)EBO Emitter-Base Breakdown Voltage IE = -10μA, IC = 0 -5.0 V
NOTE letters reversed for this parameter VEB not VBE . The O means is collector open circuited.
So this time a meter connected as Mr Davis told us would find emitter negative wrt base, reverse biased, which is what that parameter specifies - reverse withstand capability of EB junction.
It's one of those little points of technique that come from playing with real parts instead of simulated ones.
The conventions
usually work if one is rigorous.
Only way her teacher is right is if that PNP has suffered an internal Collector to Emitter short . IMHO.
datasheet at
http://www.google.com/url?sa=t&rct=...K0hKshhDb60zBONXQ&sig2=cupCh8ZbZ7737zChEnpyeQ