Thermal referral: evidence for a thermoceptive uniformity illusion without touch

Abstract

When warm thermal stimulators are placed on the ring and index fingers of one hand, and a neutral-temperature stimulator on the middle finger, all three fingers feel warm. This illusion is known as thermal referral (TR). On one interpretation, the heterogenous thermal signals are overridden by homogenous tactile signals. This cross-modal thermo-tactile interaction could reflect a process of object recognition, based on the prior that many objects are thermally homogenous. Interestingly, the illusion was reported to disappear when the middle digit was lifted off the thermal stimulator, suggesting that tactile stimulation is necessary. However, no study has investigated whether purely thermal stimulation might induce TR, without any tactile object to which temperature can be attributed. We used radiant thermal stimulation to deliver purely thermal stimuli, which either were or were not accompanied by simultaneous touch. We found identical TR effects in both the original thermo-tactile condition, and in a purely thermoceptive condition where no tactile object was present. Control experiments ruled out explanations based on poor spatial discrimination of warm signals. Our purely thermoceptive results suggest that TR could reflect low-level organization of the thermoceptive pathway, rather than a cognitive intermodal modulation based on tactile object perception.

Figure 1. Experimental set up and conditions.

(A) Thermal radiant stimuli were delivered by a 125 mm diameter, 250 watt infrared light bulb using three different stimulation intensities by connecting the bulb to one of three dimmer switches preset at 0% (no stimulation), 40% (low intensity), and 100% (high intensity). The participant’s right hand rested above the infrared source. Intermediate and distal phalanges of the index, middle and ring fingers were exposed to the thermal stimulation. In the thermo-tactile condition, the fingers rested on a sheet of glass. (B) To generate the non-uniform condition, an aluminium shade was placed between lamp and middle finger, to cast a heat shadow over the middle finger.

Figure 2. Thermographic images.

Thermal infrared imaging data recorded in a participant. A thermographic camera was used to film the entire experimental procedure. Two single frames were extracted, depicting the thermal profile of the hand immediately before and after warm radiant stimulation. A region of interest corresponding to the area of the skin exposed to the stimulation was marked on each finger. The change in temperature for each experimental condition was computed as the difference between post- and pre-stimulation mean temperature within each region of interest. Uniform pattern of stimulation (top row) induced an overall increase in temperature in all finger, with no differences between the middle finger and the outer fingers (middle: 4 °C; outer fingers: 3.8 °C, tactile conditions averaged). Conversely, the non-uniform warm-neutral-warm patter (bottom row) triggered a selective increase in temperature in the outer fingers, while the temperature of the shaded middle finger did not change (middle: 0.1 °C; outer fingers: 4.5 °C, tactile conditions averaged).

Figure 3. Thermal uniformity perception.

Sensitivity (d′) measures in the purely thermal and thermo-tactile conditions. Performance was very poor in both experimental conditions, confirming a TR effect. No significant difference was found in sensitivity values between purely thermal and thermo-tactile conditions. Error bars show SE across participants.

Figure 4. Thermal intensity perception.

Participant reported when the thermode on the forehead reached the same temperature as the overall thermal sensation across index, middle and ring fingers. Overall intensity of physically non-uniform stimulations (middle finger shade present) was judged less intense than uniform patterns. No significant difference was found between purely thermal and thermo-tactile conditions. Error bars show SE across participants.

Figure 5. Thermal localization.

Participant reported whether the thermal stimulation was delivered on the index, middle or ring finger. Overall accuracy is significantly different from chance level (indicated by a dashed line). Error bars show SE across participants.