1Since pre-historic times, humans have needed to move through the environment. Being able to find one’s way back after hunting, to memorize water or food sources, or to localize safe shelter make the difference in terms of survival, and maintaining a sense of direction (SOD) and location during navigation is a fundamental cognitive function. Spatial navigation is a complex cognitive process that requires the ability to retain the spatial layout of an environment, to memorize and recognize complex visuo-spatial configurations, to find a shortcut between two locations, or create an interconnected network among different paths¹. Many people wrongly believe that SOD is an innate capability, but cognitive theoretical models of navigation suggest otherwise. To successfully navigate, individuals have to locate themselves correctly in the environment, identify the destination goal and plan the navigational route towards the goal. It has been assumed that to carry out these steps they have to access to at least two types of spatial representation: the online representation of their position in the environment, and the offline representation of the environment². Thus, spatial navigation requires an individual to process a huge amount of information in order to create a stable mental representation of the environment, that is, the “cognitive map” of the environment³. All environmental knowledge models stress the important role of the environment and the continuous human-environment interface, as well as the flexibility of navigational strategies which have to be adapted in accordance with the environmental features to allow successful, effortless navigation. Iaria and colleagues (2007) demonstrated in an fMRI study that participants change their strategy as environmental familiarity increases, so that after the first exploration a schematic, undetailed representation of the environment is generated which is enriched with details and becomes more accurate in the successive exposures to the environment and increasing familiarization with it. Very early representation of a new environment is highly likely to be an egocentric/route representation⁴ of the sequence of movements the individual performed during the exploration, and of the sequence of views he or she perceived. This type of representation develops in an allocentric/map-like representation after the individual acquires full knowledge of the environment. Thus, environments that are well known are more likely to be represented in a survey format⁵, even if an overall survey representation may be acquired from the first exposure to an environment⁶, and people with equal levels of familiarity with an environment differ in the extent and accuracy of their spatial knowledge. However, Nori and Piccardi (2011) also found that individuals with a poor SOD may accurately perform navigational tasks when they are confident in the environment, since it is a well-known and very familiar one.

2What determines proficiency in navigation is still matter of debate, as are the reasons why some individuals develop very high levels of navigational skills, while some others need much more time to become familiar with an environment. In a seminal paper, Siegel and White (1975) posited that environmental knowledge may be acquired according to three successive phases: landmark, route and survey. In the landmark phase, individuals develop a sort of figurative memory of environmental objects, that serve as “beacons” for a salient landmark; in the route stage, they acquire paths connecting different landmarks, based on body coordinates, that is to say, using an egocentric frame of reference; finally, in the survey stage, they develop a global map-like representation of the environment which involves the encoding of directions and distances between landmarks, regardless the individual’s position, and that corresponds to an allocentric frame of reference. Siegel and White (1975) argue that the organization of spatial knowledge across different stages is cumulative: the landmark representation is characterized only by landmark properties, whereas the route representation is characterized by the features of landmarks and paths, and the survey representation includes the features of all stages. Based on such a model, different levels of proficiency in navigational skills depend on the developmental stage an individual has reached: individuals who reach the survey stage will be very good navigators, while those who never develop beyond the landmark stage will be very bad navigators. Three different navigational strategies, corresponding to the three styles, have been described, which individuals tend to adopt selectively⁷. Hence, some individuals adopt a landmark strategy, which is hardly effective, and show poor performances in navigational tasks; others adopt a route strategy, and have better performances than the previous individuals; and yet others adopt the highly evolved survey strategy, and demonstrate the best performances⁸.

3In a recent study, Piccardi et al. (2016) found that landmark, route and survey individuals differ not only in their proficiency in learning and recalling environments, but even in their visuo-perceptual exploration of new environments. The researchers recorded different patterns of eye movements during the exploration and learning of the map of a schematic and simplified environment. This suggests that the three different strategies correspond to differences in the attention paid to various environmental features (i.e., near landmarks; distant landmarks; whole configuration of landmarks), and the ability to identify the main, salient features, and therefore to memorize the more useful pattern of information.

4Spatial navigation is a very precocious skill, even if only some navigational competences are already present in very young children⁹, while others develop gradually over time¹⁰, towards the highly complex adult ability. During the first year of life, children start to develop an awareness of their own motion in space. By the age of 6-9 months, they are able to find their bearings in the environment using only egocentric strategies (turning to the left/right to reach a target)¹¹. At 11 months, they start to use information pertaining to landmarks and landmark arrays¹². Children between 18 and 24 months of age can re-orient themselves and find hidden objects using geometric environmental information (i.e., the shape of the experimental room¹³) or a combination of visual landmarks and information derived from their own movements¹⁴. Five-year-old children are able to find locations in a spatial array from a novel viewpoint using landmarks alone¹⁵. Egocentric information is gradually supplemented with allocentric coding and the relation-place strategies required for building cognitive maps that start to develop around 7 or 8 years of age and are fully functioning by the age of 10¹⁶.

5Although some skills are very precocious, the topographical memory, a specific type of memory used for processing and storing information on navigational space, develops slower with respect to verbal and visuo-spatial memory, suggesting that this competence needs to be practised to reach full development¹⁷. The ability to bind geometric environmental features, landmark identity and directional assessments develops at the age of 6-8 years, in accordance with the development of visuo-spatial processing and language¹⁸.

6Despite the increasing amount of data on healthy children that has shed some light on the development of navigational memory and the interplay between visuo-spatial language and executive functions in the development of evolved navigational skills, the determining factors of individual differences are still unknown. Even the well-known effects of gender differences (males navigate better and learn quicker than females) are not present in children¹⁹. However, in recent years a neurodevelopment disorder called Developmental Topographical Disorientation (DTD²⁰) has been described that affects the ability to develop navigational competence in spite of a normal development of other competences and the absence of brain lesions or malformations. Understanding the neurocognitive bases of DTD may provide new insights into the mechanisms underlying human spatial navigation, the developmental factors of which might contribute to determining the adult level of navigational skills. There is currently still an unknown factor that affects the brain’s ability to receive and process navigational information, causing DTD. Some authors have placed this disorder in the category of selective developmental disorders, which includes dyslexia, developmental prosopagnosia and selective language delay. They explain that individuals with DTD normally develop other cognitive abilities, but complain of a specific and pervasive topographical deficit throughout their life²¹. The first reported case²² was a 43-year-old woman (Pt1) who had no brain injury or psychiatric disease, but showed persistent difficulty in topographical orientation. As a navigational strategy, she successfully used verbal scripts that helped her in orientating herself in route-based navigation tasks. She was also able to identify landmarks in a landscape.

7However, when she had to orientate herself in unknown places, she got lost frequently, and she needs a lot of time to verbally label the environmental features. Although she showed a severe deficit in the formation of a mental map of the environment, once she had acquired such a map through overtraining, her performance on the retrieval task was similar to that of a control group²³. Bianchini et al. (2010) subsequently reported the case of F.G., a 22-year-old man showing a more pervasive disorder, including almost all processes involved in topographical knowledge and environmental navigation. Specifically, he was unable to learn the path shown by the examiner in a route-based navigation task, and to follow a path drawn on a map. He moreover presented a deficit in translating the visuo–spatial information of the maps into verbal instructions. On the other hand, he was able to reach a place following verbal instructions when someone else provided them. Unlike Pt1, he failed in segregating and identifying a landmark in a landscape, and even when he recognized a landmark, he did not know its location or the directional information he could derive from it. In general, F.G. recognized landmarks when they had a semantic meaning (e.g., Spanish Steps in Rome was the place in which the film Roman Holiday was filmed).

8Other well-documented single cases²⁴ and group studies²⁵ have been reported in recent years. Taken together, these papers show that DTD is not a homogeneous disorder since it may affect different aspects of navigational skills. What all the individuals affected by DTD share is the presence of selective impairments in navigational skills, and the absence of deficits in other cognitive domains.

9Deficits in landmark processing in the absence of any impairment in the visual processing of faces or objects have been reported in almost all of the patients, albeit with differing degrees of severity. Thus, some individuals seem unable to recognize landmarks²⁶. J.N., for instance²⁷, was reported to be unable to recognize familiar places and streets, and, as described above, F.G. recognized famous or historical buildings (for example, the Colosseum), but had no idea of their location in the city and was unable to assess the distance between two historical buildings. L.A. and Dr. Wai showed a good recognition of landmarks, being able to recognize not only famous monuments, but also buildings and architectural features belonging to their familiar environments²⁸. L.A. was however unable to put in the right sequence landmarks belonging to familiar routes, a task that Dr. Wai performed flawlessly.

10In daily life navigation, Dr. Wai relied on landmark and route strategies, as well as verbal scripts about directions (i.e., “turn right to Bernini’s fountain, go straight forward at Trajan’s Column, etc.”), but since he lacked the processing of metric knowledge about the environment, he got lost whenever he “missed” a landmark, because he was unable to recognize that he had moved too far without reaching the expected turning point²⁹.

11In summary, a perusal of single cases evidences the role that the processing of landmarks has in the development of adult navigational skills. At a very early stage, individuals learn to “identify” landmarks, that is, they learn which environmental stimuli can be used as landmarks because they are stable elements with a fixed location (for example, buildings, the parking lot, monuments, shops, etc.), and which can not (for example cars, birds, working men, etc.) because they are mobile elements whose location in the environment could easily change in a very short time. MGC³⁰, apart from being unable to derive directional information from the landmarks she recognized, also showed difficulties in identifying what environmental elements could be used as landmarks (for instance, she once relied on the position of the gardener who was pruning roses).

12A successive step consists in becoming able to “attach” spatial information to the environmental elements. A landmark is indeed a special object whose identity is the sum of its visuo-spatial appearance plus its spatial location³¹, and which is processed by special brain areas: the parahippocampal place area, or PPA³², and a temporal area of the dorsal stream, the so-called “what pathway” (responsible for visual stimuli identification), which receives direct connections from the ventral stream, the so-called “where pathway” that processes the spatial features of visual percepts³³. This area is involved in assessing whether a series of landmarks belonging to the same neighbourhood is presented in the right sequence³⁴. However, it failed to show any activation when the same task was performed by Dr. Wai³⁵, even though he was able to do so flawlessly. This suggests that Dr. Wai processed environmental objects that can be used as reference points during navigation as common objects and not as landmarks (that is, as environmental objects whose identity is defined by their spatial location and not only by their appearance).

13Functional and connectivity brain anomalies have also been described in other neuro-imaging studies. Thus, Iaria and colleagues (2009) reported weak activation of the hippocampal complex and retrosplenial cortex in Pt1, corresponding to her difficulties in generating a cognitive map of a virtual environment. Palermo et al. (2014a), describing the case of L.A., reported the absence of activation of PPA and retrosplenial cortex in tasks assessing route knowledge. Functional and connectivity alterations in the network, including hippocampus, PPA, retrosplenial cortex and frontal areas in the absence of morphological anomalies were also described by Iaria and co-workers in a group of DTD (2014), and by Kim and co-workers (2015).

14The importance of the above-described network in navigation is supported by a very recent result showing that the functional connectivity between the posterior hippocampus and the retrosplenial cortex is significantly higher in individuals with high levels of proficiency in navigational skills36. At the moment it is still unclear which factors play a significant role in determining the adult level of proficiency. Some studies suggest that individual differences in navigation are mainly due to gender differences37. However, not only are gender differences not present in all studies, but in some cases women outperform males in navigational tasks38. When we compared men and women enrolled as pilots in the Air Force with men and women with no experience as pilots matched for age and education to the pilots, we found that gender differences were present only in non-pilots; pilots had better navigational memory without any gender differences, so that both male and female pilots performed significantly better than non-pilot males39. It is possible that the high level of navigational memory in the absence of gender differences, that we observed in pilots, is due to the training they had received in the Air Force Academy; but it is also possible that women enrolled to become pilots already had high levels of navigational memory40. Further studies are needed to explore the nature of individual differences.

15In conclusion, the study of DTD has shed some light on the role that the development of landmark processing may play in the development of adult navigational competencies and the brain areas involved in the development of high-level navigational skills. It is currently clear that evolved navigational skills are related to increased connectivity in the hippocampal complex and the parahippocampal and retrosplenial areas. Future studies are however necessary to further our understanding of the role of early experiences and of specific training in developing the most advanced navigational skills and increased connectivity.