In an earlier post The Sugar Made Me Do It, I covered recent research by de Araujo, Oliveira-Maia et al. on how food, specifically sucrose, can reinforce eating by activating mid-brain dopamine circuitry, even in the absence of taste. In the accompanying editorial essay by Andrews and Horvath, this great graphic appeared, representing what is known about how eating can act on the hypothalamus and on the mesolimbic dopamine system (ventral tegmental area, nucleus accumbens, and prefrontal cortex).
Here is a much more convincing link to how eating can become appetite-driven, which previous posts on Genetics and Obesity and On the Causes of Obesity had raised as an important issue in the obesity problem.
Just one more note on the graphic: in terms of how taste can affect dopamine function, see some thoughts in the post on the neuropeptide orexin.
Figure 1. Schematic Illustration Depicting Some of the Major Findings of de Araujo and Oliveira-Maia et al
Taste alone (noncaloric sweetener), taste with caloric value (sucrose solution), or caloric value only (in the absence of taste receptors) can all equally activate the midbrain reward circuitry. To date, major emphasis has been placed on the hypothalamus and its various circuits, including orexin (ORX/Hcrt)- and melanin concentrating hormone (MCH)-producing neurons in the lateral hypothalamus as well as neuropeptide Y (NPY)/agouti-related protein (AgRP)- and -melanocyte-stimulating hormone (-MSH)-producing neurons in the arcuate nucleus, as a homeostatic center for feeding, responding to various peripheral metabolic hormones and fuels. The mesencephalic dopamine system is also targeted by peripheral hormones that affect and alter behavioral (and potentially endocrine) components of energy homeostasis. The results by de Araujo and Oliveira-Maia et al. highlight, however, that without classical hedonic signaling associated with reward-seeking behavior, the midbrain dopamine system can be entrained by caloric value arising from the periphery. While the precise signaling modality that mediates caloric value on dopamine neuronal activity remains to be deciphered, overall it is reasonable to suggest that distinction between hedonic and homeostatic regulation of feeding is redundant. DA, dopamine; GABA, γ-aminobutyric acid; Glut, glutamate.
Neuroanthropology 
Also, here’s a Nature Neuroscience Review article on “The Neural Mechanisms of Gustation: A Distributed Processing Code”: http://www.nature.com/nrn/journal/v7/n11/abs/nrn2006.html
Abstract: Whenever food is placed in the mouth, taste receptors are stimulated. Simultaneously, different types of sensory fibre that monitor several food attributes such as texture, temperature and odour are activated. Here, we evaluate taste and oral somatosensory peripheral transduction mechanisms as well as the multi-sensory integrative functions of the central pathways that support the complex sensations that we usually associate with gustation. On the basis of recent experimental data, we argue that these brain circuits make use of distributed ensemble codes that represent the sensory and post-ingestive properties of tastants.
Dopamine release is increased in cocaine abuse, gambling, playing computer games, sex, which all usually give a feeling of pleasure. When the stimulus disappears, dopamin drops resulting in depression. This is the main “neuro-chemical” mechanism of addiction with mentioned stimuli.
After long-term dopamine stimulation, the amount of dopamine nerves in the midbrain decreases, making hard for a former addict to achive pleasure with normal activities and thus tempting him in further abuse of dopamine stimulants. This may happen with cocaine abuse, and probably with long-term over-feeding.
Jan, that’s a much too mechanistic and hydraulic view of brain function, and also a much too limited view of what dopamine does (pleasure/depression). While your approach does have a gist of truth – dopamine release is associated with a wide range of activities, addicts can have a hard time achieving satisfaction/pleasure in other arenas, temptation is central to substance abuse – it locates all the explanation in one mechanism. That just doesn’t match up with the neuroscience, or with field research with substance abusers.
You might check out an early post I wrote on dopamine, and more on research on wanting/craving with addicts. Mind Hacks also covers the latest on dopamine and such, so you could head over there to do a search for what Vaughan Bell has covered.
So how can a person whos has been self medicating per se in order to reach that “high” or just to get out of a depression, subsititute the drug of choice? Like food for example?