Yes, I'd have expected to see some 38 kHz carrier signal too.
Brian Drummond has guessed a likely cause, where the phototransistor detector has too-limited bandwidth to see the carrier frequency (perhaps 36kHz - 56kHz). It is difficult to see the rise time of the oscilloscope plot, but a guess is: 20us. Besides, photo transistors have slow response time when load resistance is chosen too-high:
A phototransistor load resistor \$R_L\$ should be chosen small enough so that 40kHz is within the passband. A larger-value load resistor limits frequency response so that carrier frequency is not seen. That BPV11F phototransistor data sheet says that you should get 100kHz bandwidth with a 100 ohm \$R_L\$, but that low value might give you a rather insensitive detector.
In addition, when probing \$R_L\$ with an oscilloscope probe, you should choose X10 attenuator setting, rather than X1 attenuator. The additional capacitance of X1 probe in parallel with \$R_L\$ will reduce frequency response even more.
Your phototransistor detector circuit may have such limited bandwidth, that it serves as a crude demodulator, yielding an output that looks similar to a proper IR-remote detector chip.