Very strange analysis given in yourbook for this example problem. Giving a circuit with all the part values including the transistor beta, then imposing a particular collector current is not right. Consider that a typical npn silicon transistor is unlikely to survive an 18.5 V base-emitter voltage! It's essentially a forward biased diode, and the resulting current would be huge.
Instead, I would have started with an approximation for the base current, assuming forward bias:
##I_B = \frac{(20 - 0.7)\; V}{120\; kΩ} \approx 160 \; μA##
Then found that if the transistor was operating in its linear region that this base current, via the transistor's β, would want to produce an ##I_C## that is closer to 32 mA rather than 2.5 mA . Clearly that can't happen because 32 mA flowing through the collector resistor would produce a potential drop of about 160 V, and there's only 20 V to work with. From this you should be able to draw your own conclusions about what state the transistor will be in, and calculate an approximate value for the actual collector current.