Compositionality in visual perception.

Quilty-Dunn et al.'s wide-ranging defense of the Language of Thought Hypothesis (LoTH) argues that vision traffics in abstract, structured representational formats. We agree: Vision, like language, is compositional - just as words compose into phrases, many visual representations contain discrete constituents that combine in systematic ways. Here, we amass evidence extending this proposal, and explore its implications for how vision interfaces with the rest of the mind.

Problems and Mysteries of the Many Languages of Thought.

"What is the structure of thought?" is as central a question as any in cognitive science. A classic answer to this question has appealed to a Language of Thought (LoT). We point to emerging research from disparate branches of the field that supports the LoT hypothesis, but also uncovers diversity in LoTs across cognitive systems, stages of development, and species. Our letter formulates open research questions for cognitive science concerning the varieties of rules and representations that underwrite various LoT-based systems and how these variations can help researchers taxonomize cognitive systems.

Are icons sense data?

We argue that Hoffman, Singh, and Prakash (Psychon Bull Rev, this issue) have not made the case that "the language of space-time and physical objects is the wrong language for describing the true structure of the objective world." Further, we contend that, contrary to what Hoffman et al. claim, the perceptual icons posited by interface theory seem best taken to be sense data.

Behavioral, electrophysiological, and histopathological consequences of mild fluid-percussion injury in the rat.

Metabolic dysfunction in the relay nuclei of the rat vibrissa circuit follows traumatic brain injury (TBI). This study examined the effects of mild (1.4-1.5 atm) parasagittal fluid-percussion injury on the electrophysiology of this circuit. TBI caused significant reductions in slope and increases in latency of vibrissa-evoked field potentials 3 days after injury. Assessment of open-field swimming revealed an increase in thigmotaxis 2 days after injury. TBI caused mild selective cortical damage and limited axonal swelling at the injury site. Thus mild injury disrupts somatosensory electrophysiology and exploratory behavior.

Therapeutic effects of environmental enrichment on cognitive function and tissue integrity following severe traumatic brain injury in rats.

Postinjury environmental enrichment (EE) has been shown to alter functional and anatomical outcomes in a number of injury paradigms, including traumatic brain injury (TBI). The question of whether EE alters functional outcome following TBI in a model which produces overt histopathological consequences has not been addressed. We investigated this question using the severe, parasagittal fluid percussion injury (FPI) model. Rats (n = 7 per group, enriched and standard for behavior; n = 15 per group for histology) underwent severe (2.2-2.6 atm) FPI, with sham-operated rats (n = 7 per group, enriched and standard for behavior; n = 6 enriched, n = 3 standard for histology) serving as controls. Animals were allowed to recover for 11 days either in standard single housing or together (injured and sham) in an enriched environment consisting of a 92 x 61 x 77-cm ferret cage filled with various stimulatory objects. Consistent with earlier reports, injured animals recovering in the enriched environment showed significantly (P < 0.05) shorter latencies to find the platform in a Morris Water Maze task versus injured/standard animals on day 12 post-TBI. However, both injured groups showed significant deficits versus sham groups (P < 0.05). There were no differences between the sham/enriched and sham/standard groups. No significant group differences in swim speed were observed. At 14 days post-TBI, enriched animals had approximately twofold smaller lesion areas in regions of the cerebral cortex posterior to the injury epicenter (-4.5, -5.8, -6.8 mm relative to bregma; P < 0.05) compared to injured/standard animals. In addition, overall lesion volume for the entire injured cortical hemisphere was significantly smaller in animals recovering in the enriched environment. These results indicate that noninvasive environmental stimulation is beneficial in attenuating cognitive deficits and preserving tissue integrity in a TBI model which causes cerebral contusion and cell death.

Chronic failure in the maintenance of long-term potentiation following fluid percussion injury in the rat.

Traumatic brain injury (TBI) can produce chronic cognitive learning/memory deficits that are thought to be mediated, in part, by impaired hippocampal function. Experimentally induced TBI is associated with deficits in hippocampal synaptic plasticity (long-term potentiation, or LTP) at acute post-injury intervals but plasticity has not been examined at long-term survival periods. The present study was conducted to assess the temporal profile of LTP after injury and to evaluate the effects of injury severity on plasticity. Separate groups of rats were subjected to mild (1.1-1.4 atm), moderate (1.8-2.1 atm), or severe (2.2-2.7 atm) fluid percussion (FP) injury (or sham surgery) and processed for hippocampal electrophysiology in the first or eighth week after injury. LTP was defined as a lasting increase in field excitatory post-synaptic potential (fEPSP) slope in area CA1 following tetanic stimulation of the Schaffer collaterals. The fEPSP slope was measured for 60 min after tetanus. Assessment of LTP at the acute interval (6 days) revealed modest peak slope potentiation values (129-139%), which declined in each group (including sham) over the hour-long recording session and did not differ between groups. Eight weeks following injury, slices from all groups exhibited robust maximal potentiation (134-147%). Levels of potentiation among groups were similar at the 5-min test interval but differed significantly at the 30- and 60-min test intervals. Whereas sham slices showed stable potentiation for the entire 60-min assessment period, slices in all of the injury groups exhibited a significant decline in potentiation over this period. These experiments reveal a previously unknown effect of TBI whereby experimentally induced injury results in a chronic inability of the CA1 hippocampus to maintain synaptic plasticity. They also provide evidence that sham surgical procedures can significantly influence hippocampal physiology at the acute post-TBI intervals. The results have implications for the mechanisms underlying the impaired synaptic plasticity following TBI.

Secondary hypoxia following moderate fluid percussion brain injury in rats exacerbates sensorimotor and cognitive deficits.

Human head trauma is frequently associated with respiratory problems resulting in secondary hypoxic insult. To document the behavioral consequences of secondary hypoxia in an established model of traumatic brain injury (TBI), intubated anesthetized animals were subjected to fluid percussion (FP) injury (1.87-2.17 atm) followed by 30 min of either normoxic (TBI-NO, n = 10) or hypoxic (TBI-HY, n = 11; pO2 = 30-40 mm Hg) gas levels. Sham animals (n = 19) underwent all manipulations except for the actual trauma. Animals were tested on various sensorimotor tasks beginning 3 days after FP injury along with cognitive testing on days 22 through 29 posttrauma. The secondary hypoxic insult exacerbated the sensorimotor deficits on beam-walking compared to those animals only receiving trauma. Cognitive impairments were also observed in the TBI-HY group in the hidden platform task compared to FP injury alone. These data indicate that a secondary hypoxic insult exacerbates both sensorimotor and cognitive deficits after TBI. This study provides direct evidence that incidences of hypoxia after brain trauma may potentially result in an increase in neurological deficits for the subpopulation of head injured patients undergoing hypoxic conditions further warranting strict monitoring of these events.

Effects of combined postischemic hypothermia and delayed N-tert-butyl-alpha-pheylnitrone (PBN) administration on histopathologicaland behavioral deficits associated with transient global ischemia in rats.

Previous cerebral ischemia studies have reported the limitations of restricted periods of postischemic hypothermia in producing long-term neuroprotection. The present experiment attempts to determine whether delayed treatment with the free radical scavenger N-tert-butyl-a-phenylnitrone (PBN) is protective at 2 months following transient global forebrain ischemia, and whether additive effects can be observed when PBN is administered in combination with moderate hypothermia. For this aim rats were subjected to 10 min of two-vessel forebrain ischemia followed by (a) 3 h of postischemic normothermia (37 degrees C); (b) 3 h of postischemic hypothermia (30 degrees C); (c) normothermic procedures combined with delayed injections of PBN (100 mg/kg) on days 3, 5 and 7 post-insult; (d) postischemic hypothermia combined with delayed PBN treatment; or (e) sham procedures. Outcome measures included cognitive behavioral testing and quantitative histopathological analysis at 2 months. Postischemic PBN injections induced a systemic hypothermia (1.5 degrees C-2.0 degrees C) that lasted for 2-2.5 h. Water maze testing revealed significant performance deficits relative to shams in the normothermic ischemic group, with the postischemic hypothermia and PBN groups showing intermediate values. A significant attenuation of cognitive deficits was observed in the animal group receiving the combination postischemic hypothermia and delayed PBN treatment. Quantitative CA1 hippocampal cell counts indicated that each of the ischemia groups exhibited significantly fewer viable CA1 neurons compared to sham controls. However, in rats receiving either delayed PBN treatment or 3 h of postischemic hypothermia, significant sparing of CA1 neurons relative to the normothermic ischemia group was observed. These data indicate that hypothermia combined with PBN treatment provides long-term cognitive improvement compared to nontreatment groups. PBN-induced mild hypothermia could contribute to the neuroprotective effects of this pharmacological strategy.

Cognitive function following traumatic brain injury: effects of injury severity and recovery period in a parasagittal fluid-percussive injury model.

Previous work in this laboratory has demonstrated that rats show substantial deficits on the cued and hidden versions of the Morris water maze, as well as an apparent time-dependent recovery over a period of months, following moderate parasagittal fluid-percussion (FP) injury. However, the longitudinal nature of those studies precluded definitive statements regarding recovery because of the possible confound of practice-dependent improvements in performance. The present experiments were undertaken to address this issue and to investigate more closely the relationship between impact severity and posttraumatic learning/memory deficits, which have not been examined thoroughly in this model. Separate groups of rats were subjected to mild (1.1 to 1.4 atm), moderate (1.8 to 2.1 atm), or severe (2.2 to 2.7 atm) FP injury (or sham surgery) and tested on several water maze tasks at either 5 days or 8 weeks after injury. Moderately and severely injured animals showed impairment in acquisition of the hidden platform task at both time points. Cued platform task performance was impaired significantly in severely injured animals 8 weeks after insult. Mildly injured animals exhibited no significant deficits on either task at either time point. The results indicate that deficits on the hidden platform task are more robust than those on the cued platform task, and that performance on both tasks is dependent on injury severity. They also indicate that the learning/memory deficits in this model are relatively enduring, suggesting that the model is a reasonable one for assessing potential treatment regimens.

Exacerbation of cortical and hippocampal CA1 damage due to posttraumatic hypoxia following moderate fluid-percussion brain injury in rats.

OBJECT: Patients with head injuries often experience respiratory distress that results in a secondary hypoxic insult. The present experiment was designed to assess the histopathological consequences of a secondary hypoxic insult by using an established rodent model of traumatic brain injury (TBI). METHODS: Intubated anesthetized rats were subjected to moderate (1.94-2.18 atm) parasagittal fluid-percussion injury (FPI) to the brain. Following the TBI, the animals were maintained for 30 minutes by using either hypoxic (TBI-HY group, nine animals) or normoxic (TBI-NO, 10 animals) gas levels. Sham-operated animals also underwent all manipulations except for the FPI (sham-HY group, seven animals; and sham-NO group, seven animals). Three days after TBI the rats were killed, and quantitative histopathological evaluation was undertaken. Cortical contusion volumes were dramatically increased in the TBI-HY group compared with the TBI-NO group (p < 0.03). Qualitative assessment of cortical and subcortical structures demonstrated significant damage within the hippocampal areas, CA1 and CA2, of TBI-HY animals compared with the TBI-NO animals (both p < 0.03). There was also a significant increase in the frequency of damaged neuronal profiles within the middle and medial sectors of the CA1 hippocampus (p < 0.03) due to the hypoxic insult. CONCLUSIONS: The results of this study demonstrate that a secondary hypoxic insult following parasagittal FPI exacerbates contusion and neuronal pathological conditions. These findings emphasize the need to control for secondary hypoxic insults after experimental and human head injury.

Contribution of NMDA and non-NMDA receptors to synaptic transmission from the brachium of the inferior colliculus to the medial subdivision of the medial geniculate nucleus in the rabbit.

Previous work from this laboratory has demonstrated that monosynaptic inputs from the brachium of the inferior colliculus (BIC) to the medial subdivision of the medial geniculate nucleus (mMG) strengthen as a result of associative conditioning with an acoustic conditioned stimulus (i.e., fear conditioning). One model that has been proposed to underlie certain types of neuronal plasticity involves the recruitment of N-methyl-D-aspartic acid (NMDA)-type glutamate receptors. The purpose of the present study was to examine the relative contributions of glutamatergic NMDA and non-NMDA receptors to synaptic transmission within this pathway. Individual contributions of the specific receptor types were assessed through the use of 2-amino-5-phosphonovaleric acid (AP5), a selective NMDA receptor antagonist, and 6-cyano-5-nitroquinoxaline-2,3-dione (CNQX), a non-NMDA receptor antagonist. Bipolar stimulating electrodes were stereotaxically implanted in BIC and recording electrodes (attached to dual 32-gauge cannulae for delivery of drug) were positioned in mMG of New Zealand albino rabbits. Single pulses (150 micros, 100-350 microA) delivered to BIC resulted in short-latency (<4 ms) responses in mMG. BIC-evoked single-unit activity was recorded from mMG before, during, and at several intervals after injection of AP5, CNQX, and/or artificial cerebrospinal fluid (ACSF). Injection of either AP5 or CNQX, but not ACSF, significantly attenuated the short-latency BIC-evoked responses in the vast majority of cells tested. These findings suggest that the monosynaptic pathway from BIC to mMG is glutamatergic and that this pathway frequently employs NMDA-type receptors during electrically stimulated synaptic transmission. Due to the NMDA receptors' proposed role in plasticity (e.g., long-term potentiation), these results may have implications for understanding the mechanisms of synaptic plasticity observed at this synapse during associative learning.

Temporal and regional patterns of axonal damage following traumatic brain injury: a beta-amyloid precursor protein immunocytochemical study in rats.

Diffuse axonal injury (DAI) is an important consequence of human head trauma. This experimental investigation utilized the immunocytochemical visualization of beta-amyloid precursor protein (beta-APP) to document regional patterns of axonal injury after traumatic brain injury (TBI) and to determine the importance of injury severity on the magnitude of axonal damage. Rats underwent moderate (1.84-2.11 atm) or severe (2.38-2.52 atm) parasagittal fluid-percussion (F-P) brain injury or sham procedures. At 1, 3, 7 or 30 days after TBI, rats were perfusion-fixed and sections immunostained for the visualization of beta-APP. A regionally specific axonal response to TBI was documented after moderate F-P injury. Within the dorsolateral striatum, an early increase in beta-APP-positive axonal profiles at 24 hours (h) was followed by a significant decline at subsequent survival periods. In contrast, the frequency of reactive profiles was initially low within the thalamus, but increased significantly by day 7. Within the external capsule at the injury epicenter, numbers of immunoreactive axons increased significantly at 24 h and remained elevated throughout the subsequent survival periods. At multiple periods after TBI, selective cortical and thalamic neurons displayed increased staining of the perikarya. A significant increase in the overall frequency of beta-APP profiles was documented in the severe vs moderately injured rats at 72 h after TBI. These data indicate that parasagittal F-P brain injury (a) results in widespread axonal damage, (b) that axonal damage includes both reversible and delayed patterns, and (c) that injury severity is an important factor in determining the severity of the axonal response to TBI.

Functional relationship between the hypothalamic vigilance area and PAG vigilance area.

The vigilance reaction is characterized by a large bradycardia, a pressor response, and inspiratory apnea in anesthetized rabbits and the inhibition of movement in conscious rabbits. This affective response pattern can be elicited by electrical stimulation of the dorsolateral hypothalamus (the hypothalamic vigilance area) or the ventrolateral periaqueductal gray (the periaqueductal gray vigilance area). The present study sought to advance our understanding of the functional relationship between the hypothalamic vigilance area (HVA) and the periaqueductal gray vigilance area (PVA) by measuring the effects of transverse transections of the caudal portion of the ventrolateral PAG (vlPAG) upon the cardiovascular responses elicited from the dorsolateral hypothalamus and the rostral vlPAG. Selective transverse transections of the caudal vlPAG significantly reduced the magnitudes of the bradycardia and pressor response elicited by stimulation of the PVA rostral to the transection site, but had minimal impact on the cardiovascular responses evoked by stimulation of the HVA. These findings suggest that the cardiovascular responses elicited by stimulation of the vlPAG are mediated by a neural pathway that is parallel, at least in part, to the one that subserves the response elicited from the HVA. The results also provide support for the view that the PAG is not an essential structure in the mediation of the autonomic components of affective behaviors involving behavioral inhibition.

Hippocampally dependent and independent chronic spatial navigational deficits following parasagittal fluid percussion brain injury in the rat.

Previous reports have documented spatial navigational deficits following experimental traumatic brain injury (TBI), although the majority of the work to date has involved assessment at acute intervals following TBI, and has focused on tasks sensitive to hippocampal dysfunction. The present experiments were designed to investigate the chronic consequences of TBI, and the possible contribution of extrahippocampal dysfunction to TBI-induced spatial navigational deficits, in a moderate parasagittal fluid percussion TBI model. In Experiment 1, animals were pre-trained in a water maze, subjected to TBI or sham procedures, and re-evaluated in the water maze 48 h following the insult. Six to 8 weeks following TBI, the same animals were required to navigate to a different platform location. TBI animals exhibited significant deficits in retention of previously learned spatial information at the 48 h interval, and marginally impaired acquisition of a novel platform location during the chronic test sessions. In Experiment 2, animals were required to navigate to novel spatial locations using cued (to evaluate extrahippocampal function) as well as non-cued variants of the water maze task during the 8 week period following the insult. Injured animals exhibited deficits in both tasks which gradually diminished over the course of testing. The results of these experiments indicate that moderate TBI is accompanied by both retention and acquisition deficits, and that some of the navigational deficits observed in the water maze can be attributed to extrahippocampal damage. The possible recovery of spatial navigational ability following parasagittal TBI at moderate intensities is also discussed.

The role of nitric oxide in the pathophysiology of thromboembolic stroke in the rat.

Although nitric oxide (NO) has been shown to play an important role in the pathophysiology of cerebral ischemia, its contribution to the pathogenesis of experimentally induced thromboembolic stroke is unknown. In this study, we pharmacologically manipulated NO levels in the acute post-thrombotic stage and determined the effects on behavior and histopathology. The following drugs were used: nitro-L-arginine-methyl ester (L-NAME), a non-specific endothelial and neuronal nitric oxide synthase (eNOS and nNOS) inhibitor, 3-bromo-7-nitroindazole (7-NI), a specific inhibitor for nNOS, the NO precursor, exogenous L-arginine and the NO-donor, 3-morpholino-sydnonimine (SIN-1). Male Wistar rats (n = 76) were randomly assigned to receive vehicle or drug immediately after common carotid artery thrombosis (CCAT). Regional measurements of cortical NOS activity using the [3H]L-arginine to [3H]L-citrulline conversion assay were decreased 1 h after treatment with L-NAME and 7-NI by 50 and 65%, respectively; hippocampal NOS activity was reduced with L-NAME by 35% and with 7-NI by 65%. L-NAME significantly worsened forelimb placing as compared to other groups. 7-NI accelerated sensorimotor recovery. Water maze retention deficits were noted 48 h after CCAT and these were exacerbated by L-NAME treatment. Histopathological protection was conferred in the hippocampus by 7-NI and SIN-1; conversely, L-NAME increased neuronal injury in the contralateral cortex. L-arginine had no effect on these outcomes. In conclusion, both structural and functional consequences of CCAT can be aggravated by limiting endothelial NO production in the acutely post-thrombotic brain. In contrast, inhibition of nNOS and infusion of an NO donor has a beneficial effect on pathology.

Physiological evidence for hippocampal disinhibition resulting from activation of the median raphe.

The mechanism underlying median raphe (MR)-induced facilitation of hippocampal synaptic transmission was investigated by recording stimulus-evoked field potentials and unitary responses in urethane-anesthetized rats. Stimulation of the MR 40 ms prior to perforant path (PP) activation significantly increased the magnitude of PP-evoked granule cell population spikes (median increase = 78%) without affecting population EPSP slope. Injection of homocysteic acid into the vicinity of the MR also facilitated PP-evoked granule cell population spikes, in a dose-dependent manner. Nineteen dentate hilar units were characterized as putative interneurons on the basis of their waveform characteristics and their response to PP stimulation. Electrical activation of the MR inhibited spontaneous or PP-evoked activity in the majority (75%) of these cells; the remaining cells were unaffected. MR stimulation also inhibited spontaneous activity in a large proportion (60%) of putative interneurons in CA1. The present results provide evidence that neurons within the raphe modulate hippocampal throughput by altering discharge of non-principal cells. These data, thus, support the idea that disinhibition is a common mechanism by which extrahippocampal structures modulate information flow through the hippocampus.

Cardiorespiratory components of defense reaction elicited from paraventricular nucleus.

The present study was conducted to test the hypothesis that the paraventricular nucleus of the hypothalamus (PVN) is involved in the mediation or modulation of the cardiorespiratory components of the defense reaction (DR) in rabbits. Electrical stimulation of the PVN elicited increases in blood pressure and heart rate, hyperventilation, decreased blood flow to the visceral organs, and an increase in blood flow to the hindlimbs that was mediated by an atropine-sensitive vasodilation system. This response pattern is nearly identical to the one that is elicited by electrical stimulation of the hypothalamic defense area. In addition, the cardiomotor component of the baroreceptor reflex was observed to be suppressed during electrical stimulation of the PVN. Previous studies have shown that electrical stimulation of the hypothalamic defense area also leads to inhibition of the cardiomotor component of the baroreceptor reflex. The results of the present study provide evidence that the PVN is involved in the mediation or modulation of the defense reaction.

Chronic histopathological consequences of fluid-percussion brain injury in rats: effects of post-traumatic hypothermia.

Early outcome measures of experimental traumatic brain injury (TBI) are useful for characterizing the traumatic severity as well as for clarifying the pathomechanisms underlying patterns of neuronal vulnerability. However, it is increasingly apparent that acute outcome measures may not always be accurate predictors of chronic outcome, particularly when assessing the efficacy of potential therapeutic regimens. This study examined the chronic histopathological outcome in rats 8 weeks following fluid-percussive TBI coupled with moderate post-traumatic brain hypothermia, a protocol that provides acute neuronal protection. Animals received a moderate parasagittal percussive head injury (2.01-2.38 atm) or sham procedure followed immediately by 3 h of brain hypothermia (30 degrees C) or normothermia (37 degrees C). Eight weeks following TBI, serial tissue sections were stained with hematoxylin and eosin or immunostained for glial fibrillary acidic protein. Tissue damage, gliosis and immunoreactive astrocytes were observed in the ipsilateral thalamus, hippocampus, and in the neocortex lateral to the injury site. Within the thalamus, focal necrosis was restricted to selective thalamic nuclei. Significant hippocampal cell loss was found in the ipsilateral dentate hilar region of both TBI groups. Quantitative volume measurements revealed significant decreases in cortical, thalamic and hippocampal volume ipsilateral to the impact in both TBI groups. Lateral ventricles were substantially enlarged in the TBI-normothermia group, an effect which was significantly attenuated by post-TBI hypothermia. The attenuation of lateral ventricular dilation by post-traumatic hypothermia is indicative of chronic neuroprotection in this TBI model. These data provide new information concerning the chronic histopathological consequence of experimental TBI and the relevance of this trauma model to chronic human head injury.