Dear Dr. Wang,

Since the hippocampal circuits are not completely developed in 1 or 2 year-old children, they can not storage experience in their declarative LTM.

There is a beautiful paper by Lavenex and Lavenex published in Behav. Brain Res. that explain this phenomenon. The title is "Building hippocampal circuits to learn and remember: insights into the development of human memory".

Enjoy it!

I agree with both Jose Manuel and Joachim. But there is another reason. Cognizance, which is awareness of mental representations and processes. In the early years cognizance is very weak and global. Thus, toddlers are not aware of the limitations of their memory nor do they systematically organize memory to bypass them.

In rodents, it has been shown that this is due to both immaturity of neural circuits underlying memory as well as increased neurogenesis, which promotes forgetting. 

See attached for an elegant study by the Frankland lab on this subject 

http://www.ncbi.nlm.nih.gov/pubmed/24812394

with respect to other responders, there is plenty of evidence that infants and young children can remember events, people and objects.  They can't talk about it though. The question should, perhaps, be reformulated as - Why cannot adults recall what happened when they were children?  Then there are a number of possible answers, some cognitive and some neural. One cognitive answer is that the addressing system for autobiographical memory changes as we develop metamemoric abilities.  As adults our filing system is different - reflecting changes of interest and changes is the way we process the world. 

Vahid, yes.  I missed your post.  you might like Headed Records.  

Article Headed records: A model for memory and its failures

It depends on the person. I remember myself since age 4, being a small boy in the USSR, during WW2. I remember fire works, whenever Red Army liberated large cities during the war. But, it might be explained by abnormal conditions and events, caused by War.

If you restrict yourself to explicit LTM than up to 2 years it is due to language, so children have implicit memory but cannot yet make explicit memories. This is also due to the late development of the prefrontal cortex. As Susan said, it depends on how you measure it. A baby has a memory of the songs played to it in the womb and prefers mother's voice and native language over other voices! Specific special layouts and smells are also remembered from childhood, just hard to ask without getting intrusion errors. As you define LTM it is due to a later development of all the connections, i.e. higher association areas as storage sites for LTM. But there is also LTM for motor skills, and they are weakened even erased because the gait, posture etc changes dramatically from toddler to young childhood (four vs two-leg walking). Another aspect is: transfer from STM to LTM requires repetition or high salience/emotion. The latter is sometimes given, see war, abuse etc. The former is rarely given, every day is new with overwhelming new discoveries - from the child perspective, not from an adult perspective, i.e. we forget how much knowledge we acquired about the world during our first few years of life (folk physics, folk psychology ...) 

As a complement to Joachim's highlights : at the cognitive level too, in humans, interference is thought to be involved in forgetting during childhood. Behavioral studies describe this evolution in memory abilities across the ages, and are supported and enriched by imaging and electrophysiological datas. Mainly, growing connectivity and synaptogenesis in dentate gyrus and prefrontal cortices (from infancy to late adolescence) should participate to the age-related differences in conscious retrieval and episodic memory construction. Check the review of Bauer (2006, Trends in Cog Sci) for further details... Best.

In babies and young children 'epigenesis', i.e.  formation  of  functional connections (synapes) between  nerve cells and regions  in the brain  is not yet fully completed. Redundant connections  may disappear and new, more selective connections are established. It also explains why formation of new memories (consolidation) in networks of the maturing brain is still poor  or incomplete. This may  hold for  the formation of explicit as well as implicit memories.

UNFORTUNATELY, WHEN MOST PEOPLE SAY THE TERM "MEMORY," THEY IMMEDIATELY REFER TO DECLARATIVE MEMORY, DESPITE THE FACT THAT THERE ATR MULTIPLE MEMORY STEMS THAT INTERACT. THE ORIGINAL QUESTION SEEMS TO REFER TO THE PERIOD OF INFANTILE AMNESIA, IN WHICH CASE NOBODY CAN REMEMBER MUCH BEFORE THE FIRST 3 YEARS OF LIKE; HOWEVER, CHILDREN CLEARLY LEARN, AND THIS IS LIKELY DEPENDENT UPON REWARD-BASED, INSTRUMENTAL CATEGORY LEARNING DEPENDENT UPON VISUAL DISCRIMINATION WHICH DOES NOT REQUIRE WELL DEVELOPED HIPPOCAMPAL SYSTEMS; SIMILARLY, EVEN THE TWO DIOVISIONS OF DECLARATIVE MEMORY, SEMANTIC AND EPISODIC MEMORY, INTERACT WILL ALL-IMPORTANT PROCEDURAL MEMORY, AND WHILE IT IS OFTEN CONSIDERED IMPLICIT, IT'S SIGNIFICANCE FOR DAILY ADAPTATION IS MINIMIZED OR EVEN FORGOTTEN; ALL OF THESE MEMORY SYSTEMS (AND THIS IS MY NO MEANS A COMPLETE, RELEVANT LISTING), ARE RELEVANT AND FOLLOW DIFFERENT DEVELOPMENTAL TRAJECTORIES. NATURE'S SOLUTIONS TO COMPLEX PROBLEMS HAVE NEVER RELIED UPON SOPHISTICATED FORMS OF COGNITION (iTO, 1997). IMHO, A USEFUL POINT TO REMEMBER. I'M ATTACHING SOME REFERENCES; THERE ARE MANY, MANY MORE THAT COULD EASILY BE ADDED TO THIS ALL TOO BRIEF SAMPLE. LARGE CAPS FOR ME - NO EMPHASIS INTENDED.-LK

MORE REFERENCES -  THESE PAPERS, AND THE ONE ATTACHED ABOVE, SHOULD PROVIDE A GOOD START INTO THE INVESTIGATION FOR A COMPREHENSIVE ANSWER TO YOUR ORIGINAL QUESTION. - LK

As I read that some parts of the brain help to transform the short term to long term memory, the hypothesis is that that key brain portion that does that is under developed in kids, so they dont form long term memory to be remembered later on. However I still remember couple of events from 1.5 yrs, so I guess it depends person to person.

There should be papers on this hypothesis, or alternate ones.

Thanks Leonard for your  note on different developmental  trajectories for  episodic and  procedural memories. Indeed, development of  all memory systems in early childhood depends on a basic anatomical principle: the formation of  molecular components of the synaps between  nerve cells, a nerve and a receptor or a skeletal muscle. This occurs in concert with exuberant  branching of axonal and dendritic tree branches. My point is that these processes  provide a common neural basis of maturation of different forms of  learning (and memory) in early childhood. Explicit and implicit, episodic as well as semantic, procedural and motor learning. Which of course does not exclude different development trajectories, because different forms of  memory/learning depend on different neural networks of different complexity.

For example, episodic  learning/memory would depend on a network involving regions like temporal cortex and hippocampus, procedural memory on  a network involving motor cortex and basal ganglia. Since babies and young children have not yet reached 'full connectivity', they also lack the full capacity to store (or retrieve) information in their  cortical networks. This holds for episodic memory (e.g. poor representations of  specific events in time and place) as well as procedural  memory (e.g. poor  representations of motor programs, perceptual-motor skills).  'Poor representations' could imply either a) fragmentary or incomplete memory traces or b)  temporary or unstable memory traces that fail to translate into permanent structural changes in the brain.

However, infants do have the ablity  to recognise the identity of frequently encountered stimuli (like the famous  'face of mother'), and show a gradual perceptual narrowing (tuning) to specific features of  familiar faces when they grow older. This indicates that even infants can build  up  stable memory representations of biologically relevant events in their environment, if these events are frequently  encountered or repeated.

see also:

The Cognitive Neuroscience of memory, H. Eichenbaum. sec edition .Oxford Un. Press.2012

Neuronal man.The Biology of Mind. J.P. Changeux. Oxford Un.Press 1985.

The Oxford Handbook of Face Perception. Andrew. Calder e.a. Eds, Oxford Un. Press, 2011

Schacter DL, Moscovitch M. Infants, amnesics, and dissociable memory systems. In: Moscovitch M, editor. Infant memory. Plenum Press; New York: 1984

I JUST LEFT 6 PAPERS, THE FIRST REPRESENTING A GOOD SUMMARY OF DEVELOPMENTAL CIRCUITRY WITHIN THE HIPPOCAMPAL SYSTEM; OTHER PAPERS BUILD ON THAT. HAVE YOU HAD TIME TO READ AND DIGEST ALL 6 PAPERS? I NOTICED OUR MESSAGES ARE 3 HOURS APART. I, MYSELF, AM  NOT NEARLY SO FORTUNATE IN ABILITY.  I READ AND DIGEST THESE THINGS SLOWLY, TAKE NOTES, RE-READ; WOW! YOU ARE FAST!  YOU "GOT ME" BY MILES AND MILES! PRETTY IMPRESSIVE! -LK

Indeed, there are several reasons for childhood amnesia: a) neurological, and b) cognitive. To recall an experience it is necessary to have developed the ability to episodic memory. Episodic memory deals with information that is anchored in space and time and an autobiographical quality. Until children have not developed the ability to understand space and time concepts, and an awareness of self, it is not possible to recall autobiographical events. This does not mean that children lack other types of memory (procedural and semantic). As semantic memory develops earlier, children under three years could "know" autobiographical facts (but lack the experience to remember). This is important, because the semantic memory is constituted by non-contextual information, and therefore can not distinguish between experienced and other information from other sources. This phenomenon is related to the development of metamemory ability. Thus, younger children could "remember" things that never happened.

Indeed, preschool children "remember" things that never happened because they are not aware of the constraints that specify the limits between processes. Thus, if needed, they can fill in lags of memory by inference, because they do not "feel" that they violate anything. Insight into the constraints of different cognitive processes, that comes around the age of 6 years, protects them from this "false" or "made" memories. 

In my opinion the childhood memory is restored, however, the child cannot correlate things and comprehend the meaning of that restored memory.  Perhaps this also has to do with the time of  myelination. 

DR. EMERY,

I HAVE BEEN STRUGGLING WITH THIS ISSUE FOR SOME TIME, AND PERHAPS YOU CAN HELP. ALL OF THE "SENSORY" TYPE MEMORIES YOU REPORT ARE DOCUMENTED; IN FACT, I STRUGGLE WITH THE IDEA THAT IN SIMPLY OBSERVING CHILDREN BEFORE LANGUAGE DEVELOPS, THEY CLEAR DEMONSTRATE "MEMORY" IN BEHAVIOR. HOWEVER, IS IT POSSIBLE THAT THESE ARE NOT HIPPOCAMPAL SYSTEM AND/OR LANGUAGE DEPENDENT MEMORIES, BUT INSTEAD, INSTRUMENTAL LEARNING MEMORIES WHICH ARE BASAL GANGLIA DETERMINED, AND EVEN "CLASSICAL REINFORCEMENT" TYPES OF MEMORIES, DRIVEN AT LEAST IN PART BY CEREBELLUM? MY PROBLEM IS THIS: BASAL GANGLIA SYSTEMS AND CEREBELLAR SYSTEMS ARE BOTH RECIPROCALLY CONNECTED WITH CORTEX, AND ARE MUCH MORE MATURE AND FUNCTIONAL THAN HIPPOCAMPAL SYSTEMS, WHICH FOLLOW A COMPLETELY DIFFERENT DEVELOPMENTAL TRAJECTORY AS INDICATED IN A PREVIOUS REPLY I OFFERRED, ALONG WITH AN ATTACHMENT RE: HIPPOCAMPAL CIRCUITRY DEVELOPMENT; AND THIS POINT DOES MAKE SOME SENSE TO ME, SINCE THE OPERATIONS OF THESE SYSTEMS LIE OUTSIDE OF CONSCIOUS AWARENESS, AND AGAIN, ARE INDEPENDENT OF BOTH HIPPOCAMPAL SYSTEMS AND "LANGUAGE." I DON'T KNOW HOW TO TELL THE DIFFERENCE, BUT, THIS ALTERNATE VIEW WOULD ACCOUNT FOR INFANTILE AMNESIA. -BOLD FONT FOR MY VISUAL PURPOSES. - LK

Hi Virginia,

 I argued earlier that different  different forms of memories are  likely to  follow different development trajectories, because they depend on different  neural systems  that  vary in complexity. Emotional functions in infants  probably depend to a large extent  on  relatively simple  hardwired structures in the limbic system that mediate  innate and unconscious  forms of  emotions and  emotional memory.  Examples are  fear conditioning and startle reactions.  However,  'feelings' (consciously experienced and recollected emotions),  although they still depend on  the machinery of the limbic system, also utilize  systems in  (e.g.  frontal) cortex that are likely to develop later in the life span. It may not be  the language system per se that mediates these transitions but  rather the state of connectivity  in the developing brain.

Interestingly, a recent  fMRI study by  van den Heuvel revealed  early presence of key organizational architecture of regions and white matter connections in the neonate brain, albeit in an immature state.  This suggests that  early development of cognition  is a matter of  increase of  integrational  capacity and  maturation  of microstructures such as  pruning and growth of synapes. Seen from this angle, childhood amnesia is not a reflection of the defective ability to retrieve information from early childhood.  But rather of information that  never had the chance to become firmly established  in memory, due to the immaturity of neural connections.

The Neonatal Connectome During Preterm Brain Development. M.P. van den Heuvel et al. Cerebral Cortex,  May 2014. 1-14

I AGREE WITH YOU 110%; AND, I BELIEVE WHAT YOU SAY CAN REPRESENT A VERY GOOD ARGUMENT FOR EXPLAINING INFANTILE AMNESIA.  THANK YOU FOR THE INPUT. - LK

The children could not recall earlier childhood memories because as they were growing more and more synapses were being formed and it atkes some time before they comprehend the meaning of those memories and as the child learns the language skills, the child could verbalize the meaning of those memories.  Once they learn the language skills, the more one repeats any learned fact again and again, the synapses are repeatedly used and circuits are formed (reverberations) and then it gets into the long-term memory.

I HAVE HUGE PROBLEMS WITH STRICT LOCALIZATION THEORIES OF MEMORY BECAUSE MEMORIES ARE BOTH DISTRIBUTED AND LOCALIZED, AND  STRICT LOCALIZATIONIST THEORY FOR ANY FACET OF COGNITION GOES AGAINST IDENTIFIED PRINCIPLES OF LARGE SCALE BRAIN NETWORK SYSTEMS; ALSO, WHY DO MODELS OF ANY FACET OF COGNITION HAVE TO BE BASED UPON CORTICO-CENTRIC PROCESSES? THE EVIDENCE JUST DOESN'T SAY THAT. THERE ARE TONS OF DATA ON DIFFERENT TYPES OF CATEGORY LEARNING; IMPLICIT/EXPLICIT LERANING INTERACTIONS; INSTRUMENTAL LEARNING; SENSORIMOTOR LEARNING; "EMBODIED COGNITION." SO WHY DOES EVERYTHING HAVE TO BE BASED UPON CONSCIOUS COGNITIVE CONTROL PROCESSES? THIS GOES AGAINST THE 500 MILLION YEARS OF CONSISTENCY IN VERTEBRATE BRAIN EVOLUTION AND ORGANIZATION. - LK

Hi Yingxu Wang; It may be due to development of brain itself taking place during early years of life. Of Course language may be another problem. But as very rightly said even after fully developed minds may remember smell and taste of some thing which they can recognize but not able to explain. Regards