Abstract: At the beginning of the twentieth century, behaviourists like John B. Watson (1878-1958) changed the focus of attention from the inside of the brain (mentalism and introspection then being the main trend in psychology at the time) to the outside (Watson, 1913). They believed that we could learn nearly everything about animals and humans by studying their performance in learning experiments, and this was both measurable and verifiable. Today in the first decade of the twenty-first century, there has been a return to the inside. The neurosciences seek physiological explanations and connections between external behaviour and the neural mechanisms within the nervous system. With the revolution in magnetic resonance imaging (MRI) technology researchers are now able to visually represent neural activity. Other researchers have developed mathematical models and programs to visualise the patterns created in the periphery prior to central integration The author in this paper would like to distinguish these descriptive forms of representation from actual representations, i.e., those of which the animal is actually aware or conscious. Why does an animal sometimes make perceptual mistakes? (Case Study I “The Turtle and the Plastic Bag”). Is there more to dispositions? (Case Study II: “Taking Representation for a Walk. Argos and the Fake Daniel Dennett”). How is prey represented to an animal? (Case Study III “Representation of Prey in the Jellyfish/Herring Predator-Prey Dyad”). Does a simple animal feel pain or suffer? (Case Study IV: A Can of Worms. The Earthworm as Bait) It will be argued on the basis of contemporary biosemiotic research that animals (including both vertebrates and invertebrates) represent environmental information internally, and these representations can be subdivided into i.) primary or peripheral representation and ii.) central representation which are quantitative and qualitative respectively. Sensory information is conveyed via signals, these are received as stimuli then transduced into internal signals (see Theoretical Framework). At this stage the animal is not aware of the quality of the information as it has not yet been integrated or processed in a ganglionic complex. One can describe the properties of this pre-integrated information as quantitative and syntactical i.e., spatial and temporal ordering of incoming signals and their relations. The sign which is the smallest unit of qualitative representation arises only after integration of information from two or more discrete sensory modalities. These findings have repercussions for current models of animal learning and behaviour, especially in lower invertebrates (the principal subject of this paper); they also challenge the development of robots based on so-called simple systems