Speech recognition

Theory has it that language development is an ‘innate biological process’. First we learn to segment a stream of sound into syllables and words, and then we begin extracting the rules of syntax needed to generate sentences. What’s amazing is that exposure to speech is all that’s required. No formal training is needed ..interaction in a verbal community is sufficient. Based on this theory, a humanoid named DeeChee was created to mimic the way infants learn to recognize syllables and words. It was also tuned to boost the prominence of words signalling encouragement. Starting from scratch, DeeChee was able to learn simple words in minutes by just having a conversation with someone [ link ]. One small conversation for a robot; a canticle of possibilities for mankind..

Sound mind

It’s well known that the brain receives information from the ears in an orderly fashion. Signals for different frequencies arrive in tonotopic order along the auditory cortex. This means high tones are processed at one end of the auditory cortex while low tones are processed toward the other end. The location where these signals arrive determines our perception of pitch [ link ]. What’s interesting is that this type of organization applies to other properties of sound as well. Synapses that quickly release transmitters provide us with information about the onset of sound, like the beat, while synapses that release neurotransmitters more slowly provide information on qualities like timbre that persist over the duration of a sound. What’s new? Researchers have found that the pathways carrying these different synapse types are not grouped randomly. Instead, like orchestra musicians sitting in their own sections, they are bundled together by the property of sound that they convey. Tonotopic organization is preserved here as well. This means that beat and timbre have their own locations in the brain contributing to the perception of sound and music  [ link ].

A cure for Siri

Consider the phrase, “Man on first.” It doesn’t make much sense unless you know baseball. Or imagine a sign outside a store that reads, “Baby sale - One week only!” You easily infer that the store isn’t selling babies. Computers can’t do that. They haven’t mastered the pragmatic component of language yet .. information that is only available by knowing what social context prevails. However, Stanford psychologists have created a mathematical model that helps predict pragmatic reasoning [ link ]. This could allow computers to recognize when to apply commonly held social rules. Who knows, they may have just discovered a cure for the speech impediment suffered by Siri – a natural language interface for the Apple iPhone [ link ].

Siri don't parse

Siri is a natural language interface for the Apple iPhone 4S. It stands for Speech Interpretation and Recognition Interface and can answer your questions or make recommendations by delegating your requests to services on the web [ wiki link ]

Siri can’t parse pragmatics. The literal and rule-based contents of speech (semantics and syntax) are no problem. But the pragmatic content that’s supplied by context and culture is beyond Siri. However, roughly two-thirds of human communication requires intuiting one anothers meaning on a pragmatic basis [ thesis ]. We are pretty good at striking a balance between how specific and how general we need to be in order to keep the conversation rolling ..without either getting bogged down or sounding incomprehensible [ link ]. Computers on the other hand can’t do this. So, half the time I expect Siri to be either clueless ..or respond in gibberish.

Expectation-based perception

The journal Nature reports that expectation of a sensory event can increase the speed and accuracy we perceive it [ link ]. “Animals are not passive spectators of the sensory world in which they live. In natural conditions they often sense objects on the bases of expectations initiated by predictive cues. Expectation profoundly modulates neural activity by altering the background state of cortical networks and modulating sensory processing” [ link ].

Expectations alter perception ..I know this from my own practice. Expectations amplify and channel speech perception. Listeners have to take a moment to adjust when speakers say something that defies expectation. This study shows that the same holds true for taste. In one sense experiences arrive one by one, always fresh and new, but over time they become familiar re-enactments of prior experience. For instance I like yogurt for breakfast ..so I say it’s good and look forward to it when I get up in the morning. Good is a property I supply as something I experienced so long ago I don’t remember. It’s no longer fresh and new but an experience that my expectations enhance.

PTSD

Researchers in Europe have found that a gene, linked to improved memory performance ..also increases susceptibility to PTSD [ link ] “..in dramatic fashion, a gene variant responsible for differences in brain activity during memory encoding can also lead to increased risk of developing PTSD in response to catastrophe.” According to Neurobiologist James McGaugh at the University of California, Irvine “It’s well known that emotional arousal enhances memory consolidation, which can be a contributing factor to PTSD ..these findings provide genetic support for this hypothesis.”

Losing my religion

“This study applied a dual-processing model of [the mind] to show that analytic reasoning overrides flawed intuition and diminishes religious belief” -Gervais  [ link ].

I’m no mystic but I believe there’s a flaw in this logic. Claiming that analytic reasoning supercedes intuitive thinking is a false comparison. Analytic reasoning is not native; it is an acquired skill. It doesn’t have it’s own ‘processing system’ in the brain. The mind is a hybrid and analytic reasoning is just one of many ways of knowing. Discovery and invention come about just as often by integration and ‘insight’. Navigating unfamiliar territory is faster by seeing the relationship between vague and loosely-connected information than by using step-by-step analysis [ link ].

 

Analytic reasoning takes practice and devotion to it means neglecting other skills. When I consider all the technical writing I’ve done that required logical analysis; it’s a wonder I can suspend critical thinking long enough to accept things as they are. This may be dangerous. Invention takes a creative leap - a vision outside the confines of analytic thought. I do feel that technical writing has been stifling. For instance, after working on predicate-based software for so long (the kind that runs on legacy systems); it took me a while to grasp the more intuitive-based ‘object-software’ - which are the ‘apps’ that power devices running on the Internet today. After all, this was a sea-change brought about by the invention of Mitch Kapor and the vision of Steve Jobs. Neither of them knew for sure what the pay-off would be. It required a leap of faith.

Sensory integration

This garden universe vibrates complete. Some we get a sound so sweet. Vibrations reach on up to become light, and then thru gamma, out of sight. Between the eyes and ears there lay, the sounds of color and the light of a sigh.  ~   Moody Blues

   

A recent study shows that baboons can learn to tell the difference between word and non-word strings [ link ]. In an accompanying editorial, Dr Michael Platt says this evidence suggests dyslexia might be more of a visual problem than a problem matching sounds to letters. It’s a rare opportunity that I get to hear a neuroscientist jump to an either-or conclusion like that. Reading is an integrated process that coops many areas of the brain. A problem in the visual system can produce a reading deficit just as profound as a problem in the auditory system. These systems are integrated by higher centers of the brain during reading. What these investigators discovered with baboons is a better example of pattern recognition. It’s a process shared by any species that survives in the wild. Over time, the visual system is tuned to distinguish what’s meaningful from what isn’t by a process of ‘statistical regularity’. Statistical regularity simply means that certain signal combinations appear more frequently with meaningful objects than with non-meaningful objects. It’s no surprise that a baboon can perform this with a string of letters just as easily as it can a set of racing stripes. And I have no doubt that humans require a working visual system as well as a sound system to be able to read coherently. The written word hasn’t been around long enough to evolve an area of it’s own in the brain like it has for speech and vision.

 

The heart of interpretation

According to a recent study published in the New England Journal of Medicine, people who receive a diagnosis of cancer are 5 to 6 times more likely to suffer a heart attack or other cardiovascular complication within seven days after receiving the news (compared to people with similar backgrounds who are cancer-free). This is well before the disease or treatment has a chance to compromise health. Just hearing and interpreting information has an immediate impact on the cardiovascular system. This is a dramatic illustration of the relationship that exists between the mind and body [ link ].

 

Warning of a life-threatening event elicits conditioned fear. Conditioned fear depends on prior experience with a signal (hearing the term for an illness) paired with an adverse consequence. Once established, conditioned fear always involves blood pressure changes. In fact, practitioners of behavioral medicine can track fluctuations in blood pressure to fleeting psychological states ..in real-time. Blood pressure reliably follows the steps of a conditioned fear response. Changes can be observed when a signal arrives, while memories are retrieved, as the response is felt, and either panic or post-fear calm sets in [ link ].

Heartbreak beat

       

Heartbreak Beat~Psychedelic Furs

Most children I see with dyslexia don’t have problems with vision. What they have is a slight gap somewhere in their auditory system, or sense of hearing. They cannot translate the word they see on the page to the sound of a word in memory. Until they can do that .. it’s hard for them to recognize the word as something they’ve heard before. I have a theory. I believe that the first language we learn is the sound of our mother’s heart beat. It serves as an internal metronome ..setting the tempo that helps us follow the sound of speech. It marks the onset of phonemes and syllables, which then helps us identify words and sentences. Dr Oliver tells me that his stroke patients who lose their ability to speak can still sing. When they begin putting their words into a melody ..they can express themselves fluently. My friend Levi has MS. He uses music as a prosthesis. He my not be able to walk the same as he used to, but he can dance every bit as good. He thanks his early years spent at Grateful Dead concerts for that. What this suggests to me is that ‘keeping the beat’ may be a more fundamental process than learning to read or understand speech. So I’ve been pushing the envelope of my heartbeat theory ..to the annoyance of those around me. I’ve come to believe that the heartbeat we hear in the womb sets up an internal rhythm that helps us follow events for the rest of our lives. Without it ..everything we hear and see would simply appear as though they were coming out of nowhere ..then vanish into nowhere ..without coherence ..nothing helping us string them together.

Anti-anxiety

The complaint I most frequently hear from former pot smokers of my generation (those who came-of-age during the 60’s and 70’s) ..is that smoking marijuana makes them feel nervous. Some describe it as so intense, it borders on paranoia. Consequently they quit and never came back. In the day, little was known about cultivating for the psychoactive effects of marijuana.

The times they are a’ changing. A new breed of cultivator has revolutionized the field. They can grow sophisticated varietals with a range of psychoactive properties. “The THC component is the same ..it is the mixture of other elements that play a vital role in changing the psychoactive effect.” Two decades ago, most marijuana smokers bought whatever their dealer had. It was illegal even for research purposes. Consequently nothing was known about marijuana other than its THC content.

One of these newly found elements is a compound called Cannabidoil or CBD. It’s responsible for the calming and pain relieving effects found in medicinal marijuana. Takes the buzz off THC so to speak. CBD is also helpful in treating a range of problems, including arthritis, the side effects of chemotherapy, asthma, sleep disorders and post-traumatic stress disorder [ link ]. CBD is measured in grams. Your local dispensary should carry varieties containing different amounts of CBD .. 0.35 grams being considered ‘good’.

The bell jar

Conscious experience is a bell jar shaped by historical and environmental forces. What we call ‘ordinary experience’ is fashioned by exposure during early childhood and a sensory-network that has evolved over eons of earthly habitation. Recent findings published in the journal Proceeding of the Academy of Sciences help support my bell jar theory. They show that the sensory world is tuned by exposure early in life. Children growing up in cultures where their diets don’t offer much in the way of sugar, eventually lose the taste-receptor for it [ link ]. In the same way, children growing up in cultures where their language doesn’t include sounds found in other cultures eventually lose the receptor sites for them [ link ]. The neuro-pathways for discerning phonemes that are not available in the child’s early environment get ‘pruned away’. For example, children in Asian cultures that don’t have phonemes for the English ‘L’ and ‘R’ lose the ability to hear any difference between them.

Rhythm of the senses

The way information flows through the brain mediates the way we perceive the world. And an important property of information transmission is rhythm. Like a metronome, nerve cells in the brain display rhythmic activity. If one area changes it’s beat, in response to variations in external conditions, other areas adapt by changing their beat as well. This helps explains how perception changes faster than nerve signals change pathways [ link ].

Message of a thousand years

A study, published in the journal PLoS Computational Biology, suggests our brains subconsciously use a simplistic strategy in order to filter out options when faced with a complex decision. However, the research highlights how this strategy can lead to poor choices [ link ].

Message: “..a thousand years ago a massive earthquake and tsunami had all but wiped out Murohama, a fishing village on an island off the coast of Japan. The residents, knowing they were going to be inundated, sought safety on the village's closest hill. But they had entered a trap. A second wave, which had reached the interior of the island through an inlet, was speeding over the rice paddies from the opposite direction. The waves collided at the hill and killed those who had taken refuge there. To signify their grief and to advise future generations, the survivors erected a shrine ..a simple clearing by the side of a hillside road, with stone tablets and roughly made figures.”

This shrine continues to serve as a warning of where not to go in the event of an earthquake. On March 11, 2011 ..the locals heard the echoes of that message. “We all know the story about the two tsunami waves that collided at the shrine.” Instead of taking refuge on the closest hill, the one with the shrine, they took the time to get to high ground farther away. From the safety of their vantage point they saw two tsunami waves colliding at the hill with the shrine, as they did long ago. A message sent by ancestors over 1,000 years ago, traveled the distance and saved their children.

Going native

Economists tell me I make optimal decisions based on the best information available. They call this the ‘rational’ model of man. Neuroscience tells me my perception is selective, memory is faulty and decisions are mostly visceral. They call this the ‘somatic’ model of man. I think I’ll go with that. Without instinct, I’d never finish deciding what to eat or what to wear. I buy clothes that’ll make me look cool and acceptable to the people I hang with. I look for approval from my mother/dietician/checkout-girl whenever I buy groceries. I blow them off when I’m feeling rebellious. I make financial decisions fearing I’ll disappoint my father. I only bookmark pages I feel comfortable putting on facebook. I practice the ‘disarming smile’ out of fear of looking like a menace. I practice yoga to improve my posture and not come across looking like a slouch. I chose my words very carefully out of fear of looking foolish. Damn right my decisions are visceral. Appetite and desire for social approval guide my decisions just as much as any contributions from the rational side.

Attribution theory

Attribution theory says that observers tend to attribute an individual’s actions to enduring character traits whereas the individual tends to explain his or her own actions in terms of the passing demands of a situation [link]. Here’s a recent case in point. Dr. Andrew Bourne was a famous surgeon in Mammoth Lakes, California. He saved the lives of many people who’d been seriously injured on the ski slopes. In December he was charged with sexual misconduct. In January he committed suicide. Some members of the community attribute his suicide to the same ‘self-destructive’ tendencies that lead to his misconduct. Others describe him as “..energetic, optimistic and precise but perennially late” leaving the impression he had ‘contradictory tendencies’ that could help explain his behavior. According to his attorney, Dr Bourne attributed his conduct to “..a single temptation.”

Ecology of mind

Learning context  (quality of avocados increases from left to right)

At Trader Joes  [ (others)  (haas) ]
At Scolaris [ (haas)  (mex)  (others) ] 

Context dependent learning is better suited to learning in natural settings than artificial ones. Learning the quality of different avocados by sampling from selections at different store-locations improves decision making in the wild where choices come up one after another. Context dependent learning is less suited to decision making in a warehouse where choices are presented all at once in a single location. It's the difference between seeing [(haas) (mex) (others) ] in one location as opposed to seeing them appear one at a time in different locations:  

[ (haas) ]        where you’ve learned that mex generally beats haas

[ (mex) ]               but in some instances other varieties are better

[ ( others) ]            ..so it pays to keep looking.

Economists say that humans deviate from optimal choice when making context-dependent choices. On Wall Street or in a classroom, context information can be misleading. Economists call this the “more-is-less effect”. I’d say they lack ecological perspective. There’s a reason humans are sensitive to context. Psychologists are finding out that it has adaptive value in nature that you don’t see in a classroom or trading floor. It helps people make ‘optimal choices’ about which trail to take and what foods to eat in the wild. In an experiment using a species of birds called starlings, researchers at Oxford found that even though context learning may hinder performance in simultaneous prey choices; it improves performance in sequential prey encounters where subjects could reject opportunities in order to search the background. Because sequential prey encounters are more likely in nature, storing and using contextual information has greater ecological value than economists give it credit for.

Freidin et al. Science 18 November 2011 [link]

The development of appetite

It’s well known that the ability to perceive the sounds of language is tuned in childhood as a result of the interaction between exposure and early development. Infants start out with equipotential for languages. However, by the end of the first year ..neural-pathways responsible for discerning the sounds of their native language get established. Turns out that appetite for particular foods are formed this way a well, perhaps even earlier inside the womb. Infants start out indifferent to tastes, however “..later exposure changes the way taste signals are carried to the brain.” Recent studies show how exposure to a diet of crackers, cereal and bread during early development results in a lifelong preference for starchy-salt rich food [link].

Inference making

“Your job as a reader is to use your imagination and analytical skills where the author has left off.”

Intentional fallacy: it’s not what an author means to say that’s important ..it’s how the reader interprets what they say. What they intended is subject to interpretation, which isn’t necessarily going to turn out the same. But if we’re the readers, our interpretation is what matters. Communication is mostly an interpretive process. We add our perspective and ingenuity to whatever we hear or read. Attempts by the writer to narrow it down are futile ..or sterile [link]. In Harry Potter, some may see Dumbledore as gay; others might view him as quirky and without a particular sexual identity. I'm reminded of the ghost in “Hamlet” and how little we really know about him. Is he the spirit of his murdered father asking to be avenged ..? Is he a hellish apparition sent to make Hamlet commit murder ..? Or, is he just a figment of Hamlet's imagination ..? And who really gives a shit now what Shakespeare meant ..?

Immateriality

“Each person is capable of perceiving infinitely more. The universe is funneled through the reducing valve of the brain and nervous system ..what comes out at the other end is a measly trickle.”

From ‘The Doors of Perception’ by Aldous Huxley 

Only 20% of available ‘matter’ - at home or in the universe - is observable by the senses. This includes the composition of our bodies as well as all it’s surroundings. The remaining 80% is not even visible using the most sophisticated instruments of science. It’s a mystery supplied by indirection and the divinity of inference. What does this mean ..? Ordinary reality represents only a fraction of the energy that exists in the universe. The forces at work in my life are largely invisible. Perhaps the limits to what I can see are not so much physical as they are mental, like Huxley said.

Neuropharmacology of ketamines

Glutamic acid is the most abundant excitatory neurotransmitter in the nervous system. When released, it binds with the NMDA-receptor site to produce long-term potentiation (LTP).

LTP improves synaptic transmission by increasing the cell’s sensitivity to incoming signals, which allows for neuro-plasticity, learning and memory. 

NMDA (N-Methyl-D-aspartic acid) is a synthetic amino acid, which acts as an agonist at the NMDA-receptor site .. mimicking the action of glutamic acid (excitatory). 

Ketamine is an NMDA-receptor antagonist. It blocks the action of glutamic acid and acts as an analgesic at low doses. At high doses, ketamine produces a disassociative state, characterized by a sense of detachment from the physical body (depersonalization) and the external world (derealization). Users may experience what is called the “K-hole”, a period of dissociation and intense hallucinations where they experience other worlds or celestial-like dimensions ..while being completely unaware of their individual identities or the external world. Users have reported flying .. connecting to other users and objects in the cosmos ..and sharing hallucinations and thoughts with adjacent users. They feel as though their perceptions are located so deep inside the mind that the real world seems distant (hence the use of a “hole” to describe the experience).

Memory: Users do not remember the experience after regaining consciousness, in the same way that a person may forget a dream. Owing to the role of the NMDA receptor in long-term potentiation, this may be due to disturbances in memory formation. The ‘re-integration” process is slow, and the user gradually becomes aware of surroundings. At first, users may not remember their own names, or even know that they are human, or what that means. They may not be aware they have a body at all.  Ketamine is also used with local anesthetics for its amnesia action. It effectively wipes out memory for the trauma associated with severe injury and surgery.

Therapeutic value of hallucinogens

Recent studies reported in the Archives of General Psychiatry [link] and the Journal of Psychopharmacology [link], reveal that hallucinogens actually have legitimate therapeutic value. Scientists now believe these agents have the potential to help patients with post-traumatic stress, drug and alcohol dependence, unremitting pain, depression and the existential anxiety of terminal illness. According to Roland Griffiths, author of the first study:

“The psilocybin experience takes away the veil of fear and enables patients to see things in a more expanded and interconnected way. It can relieve the existential anxiety of terminal illness. The psychological improvements have helped many to reverse the course of their illness, which reinforces the notion that one should never underestimate the healing power of the psyche. Scientifically, these compounds are way too important not to study.”

This sounds familiar. In college I wrote my undergraduate thesis on the Neurological Basis of Hallucinatory Experience. I had the foresight (or audacity) at the time to recommend that hallucinogens would be a useful method-of-investigation for Psychologists. I said: “.. it would be negligent not to consider the guided peyote session as portal into alternative states of consciousness” [link].

Synthesizing minds

At a graduate seminar many years ago, a fellow student named David Stoltenberg proposed a theory that said that the simple act of reading is a “cross-sensory” event in the brain ..he even had a name for it ..“sensory synesthesia” ..which he described as “perceiving the sound of a color ..or the light of a sigh.” He was giving a multi-media presentation to demonstrate this idea ..but it didn’t turn out the way he planned ..the projectors malfunctioned ..the main point got lost ..and what I was able to get out of it left me feeling unconvinced ..it sounded too much like science fiction. When I think back, I realize I owe Dave a big apology ..and a pound of red Lebanese ..he was right ..you have to be able to “hear” what you “see” in order “understand” what you “read”.

Research now shows that synesthesia, far from being a “fringe” phenomenon, can actually enhance cognitive function in addition to being part of the reading process. Many notable artists, poets and novelist are thought to have this ability. The condition occurs from increased communication between sensory areas of the brain [link]. It probably lies on a spectrum of the way we normally perceive and experience the world. In other words, we all have it ..just some more than others [link].

Calculating minds

Success at math is often more about focusing attention and screening distractions (caused by threat and anxiety) ..than it is about activating areas of the brain actually involved with math calculation. Sian Beilock (University of Chicago) reports: 

“We used functional magnetic resonance imaging to separate anticipatory neural activity from what’s occurring while performing math. Increased activity in frontoparietal regions of the cortex, involved with focusing attention and suppressing anxiety, were better predictors of math scores than activity in regions associated with arithmetic calculation (the left intraparietal sulcus of the cortex) [link].” 

Think about walking across a suspension bridge if you're afraid of heights versus if you're not – it’s a completely different ballgame. This work suggests that educational intervention emphasizing anxiety-reduction (rather than additional math training) will be most effective in revealing a population of mathematically competent individuals, who might otherwise go undiscovered.

Parsing Gabrielle

Notes made while watching interview with Diane Sawyer.

Her speech centers are still intact ..but some of the pathways that connect speech with concepts may have been severed. They show her a picture of a table and she comes up with words all right ..just not the right ones. She’s guessing and her therapy involves prompting her to narrow down the range of possibilities until she’s in the vicinity of ‘table-ness’. It is geared toward building alternate pathways to replace the one’s she lost. The connection between her lexicon (the place where words are stored) and semantic memory (memory for meaning) may be all that’s affected. Prognosis is good. She can read words from her lexicon OK. Her difficulty is connecting them with ideas in the mind. So it’s just a process of generating alternate pathways. I wonder if she can write or type in complete sentences. I wonder if there’s a way to prompt the language pathways of the brain to act with equipotentiality, same as they did during childhood, to help facilitate the regenerative process. Apparently music can help because it activates greater brain-area ..and she can sing the words she has difficulty coming up with on her own.  Spontaneously however, she doesn’t speak in full sentences yet. Her two word utterances show a ‘return to the kernal’ ..meaning she can express the main idea without the generating the phrase-structure necessary to produce a full sentence. Hopefully, she hasn’t lost the rules of grammar ..only the ability to pick-out the words to express them. 

Kernal: When asked if she wants to return to Congress, she relies: “No, better!”
Generative grammar: Two embedded verb phrases are required to turn the kernal “No, better!” into the sentence: “No, I want to get better first”