Separate transmit and receive antennas aren't used in WiFi... because IEEE 802.11 (which the WiFi alliance uses), is a "half-duplex" protocol. That is, 802.11 cannot, by design intent, transmit and receive simultaneously.
Now with MIMO (options in 802.11n), there can be simultaneous transmissions but antennas for MIMO are typically internal and, like basic 802.11 (as above), the antennas are switched between transmit and receive.
So, the simple answer is that there's no need for separate transmit and receive antennas. It's essentially the same as using a handheld walkie-talkie radio - one antenna because it's a half-duplex communications method (the push-to-talk). In WiFi, to simplify, the push-to-talk is automated.
With other than the fanciest new MIMO 802.11n, there are often two external antennas. These are used in "switched diversity" - where the signal strength for each incoming data frame is measured in a few bits of the frame header for each antenna. The switch is made in a fraction of a microsecond. Then the "best" antenna is used for the rest of that frame. Later, a frame from other client device may be better on a different antenna.
The technical term is spatial diversity - with the basis being that in non-line-of-sight conditions, multipath and reflections may cause one antenna separated by a several wavelengths from the other, to "see" a quite different delay and reflection situation than the other. In high-end MIMO, they add Frequency Diversity - same concept, but the idea is that the conditions on a different channel (frequency) may differ. Then, to add multi-stream MIMO, the diversity ideas are applied concurrently, without switching antennas or frequencies - then there is a "vote" in a processor that chooses the best stream received vs. time, to get the lowest bit error rate.
Clear as mud?