Media experiences play an increasingly central role in the life of modern people, to the detriment of experiences in the real world. This raises the question of whether these experiences are equivalent with respect to the mental representation formed from the real-world experience. Triangulating from contradictory theories of transactive and autobiographical memory that make contrasting predictions on the effects of media (sometimes found to improve and sometimes worsen memory); in this study media is hypothesized to frame the experiences and thus to affect mental representation. The more immersive the media experience, the more informative and linguistically coherent is the resulting retelling of the experience expected to be. The language output is used to access the mental representation that in turn is based on memory. Route directions were chosen as an adequate task since it is a linguistically well described, not emotionally charged but rather instrumentalized genre that permits the control of previous knowledge. Along the route there are critical spots (strong bends, blockages, many possibilities for further advancing) where it is advisable to re-orient the imaginary walker to ensure that he stays on the planed route. In route directions these critical spots are typically marked by placing landmarks: Choosing a visible and recognizable object of manageable size to anchor the localization or movement to specific coordinates in the environment. A database of 88 route directions was collected empirically by means of an experiment. The same route was presented to adult German native speakers (22 participants per condition) that gave written route directions in German after being exposed to different randomly assigned media conditions (manipulated independent variable). The tested media were a video from a first-person perspective, an aerial plan of the route complemented with still images of the route, and a combination of both with a moving dot in the plan showing the current position in the video (multimedia condition). The baseline was provided by the prototypical experience of walking the route in the real world (no-media condition). The following hypothesis were tested: Firstly that texts composed after media-mediated experiences will not be as successful in preventing the hearer from getting lost on the way; secondly, that it is more difficult for the speaker to recognize these critical spots when relying on perspective-fixing, not self-paced media in contrast to experiencing the route in the real world (prototypical experience used as a control). Lastly, that more realistic experiences will produce more informative and coherent texts. Results of the randomized experiment support the hypotheses in part. The placement of landmarks at the different segments of the route was operationalized to account for the conceptualization of the surroundings. Results show that the plan condition is less likely to successfully lead the hearer to the final goal. The video and multimedia conditions are as well suited as the real experience to form a mental representation of the route. Blockages along the route are more difficult to identify based on an aerial view. The plan condition also results in less informative texts (fewer landmarks and fewer specifications), but not less coherent texts than in other conditions. The informative content and linguistic coherence of texts produced after exposure to the video and multimedia conditions can even outperform the control (real world condition). This is interpreted as a positive effect on memory from immersive media being used repeatedly. In the multimedia condition the drawbacks from the field-of-view of the video are counterbalanced by combining it with another media that uses an aerial perspective (the plan), resulting in an even more comprehensive experience than possible in the real world. The findings suggest that different experiences lead to different cognitive maps, that in turn lead to different texts via different conceptualizations. Apparently, the more natural the experience, the longer the produced texts and the more landmarks are used in the texts. This work contributes to an emergent literature on the role of media in shaping our representation and understanding of reality.