Umbrellas are for the future. Most of the time it is not raining, so umbrellas are useless and just take up space. They only become useful when it rains, and this will happen in the future. Other objects are also like this. Band-Aids only become useful when someone gets cut, fire-extinguishers only become useful when something accidentally catches on fire, and so on. These future-oriented objects prepare us for rainy weather, cuts, and fires that may arise in the future (and do not yet exist in the surrounding environment). Hence, temporal reasoning is required to know that one should own and keep the artifacts needed to deal with these future situations. An agent unable to anticipate future events would not bother owning an umbrella or any other future-oriented possessions.

The observation that some objects are future-oriented was inspired by Hoerl & McCormack's (H&M's) claim that tool-saving tasks (Kabadayi & Osvath Reference Kabadayi and Osvath2017; Mulcahy & Call Reference Mulcahy and Call2006) do not assess temporal reasoning. In these tasks, an agent appears to show future-thinking when an agent appears to show future-thinking when selecting a tool for use at another location (e.g., picking up a rock in one location so that it can later be used to obtain food elsewhere). H&M convincingly suggest that tool saving does not require temporal reasoning, and may instead depend on the atemporal understanding that the saved tool is useful somewhere else in the current environment. This analysis, however, led us to reflect that for future-oriented objects, like umbrellas, tool saving does require temporal reasoning.

Examining tool saving and use of future-oriented objects may help provide evidence of temporal reasoning in populations where this ability remains uncertain, such as preschoolers. As H&M note, temporal reasoning is well established in 5-year-olds, but the abilities of 3-year-olds remain in doubt (Atance et al. Reference Atance, Louw and Clayton2015; McCormack & Atance Reference McCormack and Atance2011). Indeed, 3-year-olds consistently fail tasks that require them to select an object in anticipation of a future need, such as needing a box to reach a game table (Russell et al. Reference Russell, Alexis and Clayton2010), a key to open a box in another room (Suddendorf et al. Reference Suddendorf, Nielsen and von Gehlen2011), and an item to alleviate boredom during an otherwise uneventful period (Atance et al. Reference Atance, Louw and Clayton2015; Reference Atance, Metcalf and Thiessen2017; Metcalf & Atance Reference Metcalf and Atance2011; Suddendorf & Busby Reference Suddendorf and Busby2005).

However, we see a different pattern of results when 3-year-olds are asked about future-oriented objects, though no study has done so exclusively. In one study, Atance and O'Neill (Reference Atance and O'Neill2005) asked children to select objects to take on a future trip, and to justify their selections. Although 3-year-olds referenced the future in only 37% of their explanations overall, they did so 70% of the time when justifying their selection of Band-Aids – the only future-oriented object included. In another study, Atance & Meltzoff (Reference Atance and Meltzoff2005) asked children to select appropriate items to take on imaginary future excursions. Again, Band-Aids were included as a possible item to take for a walk through a rocky stream, and 3-year-olds selected this object more often than chance and as often as older children (though they did so only when a semantically related distractor item was not present). When asked to justify this selection, 3- to 5-year-olds referenced the future in 75% of their justifications. Though these findings are encouraging, they are admittedly narrow in scope and must be interpreted cautiously. A study exclusively exploring children's beliefs about future-oriented objects would provide us with deeper insight into their temporal cognition, as a full understanding of the function of such objects necessarily involves consideration of the future.

Future-oriented objects may also be relevant for attempts to infer the cognitive abilities behind prehistoric artifacts (i.e., cognitive archaeology [Coolidge & Wynn Reference Coolidge and Wynn2016]). Scholars have suggested that even the makers and users of Oldowan tools, which date back more than 2.5 million years, were capable of temporal reasoning and could anticipate future needs (Osvath & Gärdenfors Reference Osvath and Gärdenfors2005; Shick Reference Shick1987; Suddendorf & Corballis Reference Suddendorf and Corballis2007a; see Toth & Schick Reference Toth and Schick2018 for an overview). These claims often concern tool saving. For example, Oldowan hominins transported stone tools far from the locations where they were created, transported stones and other materials to accumulation sites (which may have served as caches), and may have carried stones to hunt prey they encountered. H&M's analysis suggests that these forms of tool saving might not have required temporal reasoning. For example, rather than carrying tools for use in the future, Oldowan hominins may have viewed themselves as carrying tools for use elsewhere. But this conclusion may hinge on whether some Oldowan tools were future-oriented objects. A stone for throwing at prey is not a future-oriented object if the hunter knows the prey is already somewhere else in the current environment. But a tool specialized for butchering carcasses may be future oriented (assuming it is intended for hunted, rather than scavenged, animals). The maker may have to anticipate that an animal must be successfully hunted before the tool can be used.

We have suggested that temporal reasoning likely underlies keeping possessions that prepare one for future events. Other kinds of saving may also require temporal reasoning. Consider the act of saving raw food to be cooked. Great apes prefer many types of food when they are cooked rather than raw (Wobber et al. Reference Wobber, Hare and Wrangham2008), and given experience with cooking, chimpanzees will bring raw food to a cooking device and wait for it to be cooked (Warneken & Rosati Reference Warneken and Rosati2015). Chimpanzees are unlikely to view these activities as necessary for securing a reward that is currently somewhere else in its environment. Before the food is cooked, the reward does not yet exist. Hence, saving raw food probably indicates the ability to anticipate the future situation of cooked food being available (see Beran et al. Reference Beran, Hopper, de Waal, Sayers and Brosnan2016 for an alternative associative-reasoning explanation). Anticipating that objects will change in the future is also necessary to understand many other aspects of the world, including agriculture (e.g., seeds will grow into plants) and crafting (e.g., clay will harden, paint will dry).