Two systems come online in the first two years that will shape the rest of the lifespan: the capacity for language, which turns a wordless infant into a conversational partner in roughly thirty-six months, and the capacity for attachment, the emotional bond that becomes a template for every close relationship to follow. Both look effortless from the outside and are anything but; both have been the site of the field's fiercest nature–nurture arguments; and both, on close inspection, resolve into the same modern answer — biology prepares the ground, and social interaction does the building.
Language · the modern takeIt's not Chomsky vs. Skinner anymore
Language acquisition is still often taught as a duel between two mid-century giants. On one side stood the behaviorist account, which held that children acquire speech the way they acquire any operant behavior — through imitation, reinforcement, and the gradual shaping of verbal responses by an approving environment (Skinner, 1957). On the other stood a nativist reply arguing that reinforcement is far too weak and too slow to explain the speed and creativity of early grammar; children routinely produce sentences they have never heard and could never have been reinforced for, and the speech they hear is too fragmentary to contain the rules they end up mastering — the famous poverty of the stimulus (Chomsky, 1959). The nativist solution was an innate Language Acquisition Device (LAD), a species-specific mental faculty pre-loaded with a universal grammar whose parameters the surrounding language merely sets.
Each side captured something real. Reinforcement and imitation plainly matter — a child learns the particular words and constructions of the language around it, not some other one — yet no parent ever explicitly teaches grammar, and children raised in radically different cultures pass through strikingly similar milestones on a similar timetable, which is hard to square with pure learning. But the strong nativist claim has aged less well than its confidence suggested. The third position that now dominates the field keeps the insight that the human mind is exquisitely prepared for language while dropping the assumption that preparation requires a dedicated grammar organ separate from every other cognitive ability.
Michael Tomasello and others argue children don't need a dedicated grammar organ. Instead, powerful general abilities — pattern-detection, statistical learning, and "joint attention" (reading what a caregiver is pointing at and intending) — let kids extract grammar from the torrent of speech around them. Infants are startlingly good statistical learners: by 8 months they track which syllables tend to follow which, and use those transition probabilities to find word boundaries (Saffran et al., 1996). The consensus now: language is biologically prepared but socially built — nature loads the gun, interaction pulls the trigger (Tomasello, 2003).
Genie and Victor get taught as proof of a hard critical period — a window that slams shut. The more accurate word is sensitive period: learning is dramatically easier early and gets harder, but not impossible, later. Evidence: Genie learned vocabulary but never normal grammar (consistent with a window for syntax), while second-language accent shows a steep-but-graded decline with age rather than a cliff. And beware over-reading Genie — she suffered extreme abuse and malnutrition, so her language deficits are confounded, not a clean natural experiment.
Language · the sequenceFrom cooing to grammar in thirty-six months
Long before the first word, infants are already working on language, and the earliest signs show up in the vocal tract rather than the vocabulary. Newborn crying gives way around two months to cooing — drawn-out vowel sounds produced in states of contentment — and by roughly four to six months to babbling, the rhythmic consonant–vowel strings (bababa, dadada) that appear on a similar schedule in every hearing community and even, in manual form, among deaf infants exposed to sign language. Babbling begins universal, containing sounds drawn from many languages, and then drifts toward the phonemes of the ambient tongue, so that by the end of the first year a baby's babble already carries the accent of the language it is about to speak.
Meaning arrives around the first birthday with the first true words, typically labels for the people, objects, and routines a child cares most about. Early single words often carry the load of a whole sentence — a child who says milk may mean I want milk, there is the milk, or the milk spilled — a compression developmentalists call the holophrase. For several months vocabulary accumulates slowly; then, somewhere in the second year, many children hit a naming explosion, adding new words at a rate that can reach several a day, powered in part by fast mapping, the ability to attach a rough meaning to a novel word after only one or two exposures.
The next threshold is combination. Around eighteen to twenty-four months children begin stringing two words together — more milk, doggie gone, Mommy sock — in what is called telegraphic speech because, like an old telegram, it keeps the high-information content words and drops the grammatical connective tissue of articles, auxiliaries, and inflections. Crucially, even these two-word utterances honor word order, signaling that grammar, not just vocabulary, is coming online. Over the third year the connective tissue fills in and sentences lengthen and elaborate with remarkable speed.
The most revealing errors arrive with that grammatical explosion. Toddlers who once correctly said went and feet — memorized as whole forms — begin saying goed and foots. This overregularization looks like a step backward but is in fact a leap forward: the child has extracted a genuine rule (add -ed for the past, add -s for the plural) and is now over-applying it, even to the irregular words it used to get right. The definitive demonstration that children abstract rules rather than memorize forms is the wug test: shown a novel bird-like creature and told this is a wug, then asked about two of them, preschoolers readily supply wugs — a plural they cannot have heard before, revealing productive command of the morphological rule (Berko, 1958).
Language · perceptual narrowingTuning the ear: becoming a native listener
Speech perception is where the "biologically prepared, socially built" formula shows most vividly, because here development runs in a counterintuitive direction — infants begin as universal listeners and become specialists, losing discriminations they were born able to make. The starting point is striking: even one-month-olds perceive speech sounds categorically, hearing the acoustic continuum between ba and pa not as a smooth gradient but as two sharply divided categories, exactly as adults do and along the same phonetic boundaries their language uses (Eimas et al., 1971). Newborns arrive already biased toward speech, and having listened to the muffled prosody of their mother's voice through the womb, they prefer that voice to a stranger's within days of birth (DeCasper & Fifer, 1980).
What changes across the first year is not the ability to hear but which contrasts the system bothers to keep. In a now-classic study, English-learning six- to eight-month-olds could readily discriminate consonant contrasts used in Hindi and in Nthlakampx (a Salish language) that adult English speakers cannot hear at all; by ten to twelve months the same infants had lost the non-native distinctions while retaining the native ones — a perceptual reorganization driven by nothing more than exposure to the surrounding language (Werker & Tees, 1984). The narrowing is early and specific: exposure reshapes vowel perception by six months of age, before most infants understand a single word, so it cannot be a by-product of learning meanings (Kuhl et al., 1992).
The mechanism is best captured by Kuhl's native language magnet account, in which repeated experience with a language's sounds warps the infant's perceptual space so that variants near a familiar prototype get pulled together and heard as "the same," while the boundaries between native categories sharpen (Kuhl, 2004). Two forces do this reshaping with no explicit teaching. One is the statistical learning the update box above describes — sensitivity to how often particular sounds and syllables co-occur, which lets infants find both phoneme categories and word boundaries in an unbroken stream of speech (Saffran et al., 1996). The other is the social context of hearing: live interaction with a responsive person, complete with the exaggerated pitch and clarified vowels of infant-directed speech, produces phonetic learning that the identical audio delivered by video or over a speaker does not, underscoring that this tuning is not mere acoustic bookkeeping but a socially gated process (Kuhl, 2004).
Attachment · the theoryWhy babies bond: Bowlby's ethological turn
The dominant mid-century explanation for the infant–caregiver bond was the cupboard-love theory: the baby learns to love whoever delivers food, so attachment is a secondary drive derived from hunger. Two lines of evidence dismantled it. Ethologists studying other species had shown that the young of many animals bond through imprinting — following and attaching to the first moving thing they encounter, quite independent of feeding, as in Lorenz's famous goslings — hinting that bonding is a built-in behavioral system rather than a learned association. And Harlow's rhesus infants, offered a bare wire "mother" that dispensed milk and a soft cloth "mother" that did not, spent their time clinging to the cloth figure and fled to it when frightened, choosing contact comfort over the food source and refuting cupboard love directly (Harlow, 1958).
Weaving these threads together with psychoanalysis, control-systems theory, and evolutionary biology, Bowlby proposed that attachment is an evolved behavioral system whose biological function is protection: an infant who keeps a protective adult close is likelier to survive predators and other dangers, so natural selection favored infants who signal, seek proximity, and protest separation (Bowlby, 1969). On this view the caregiver plays two linked roles — a secure base from which the child ventures out to explore and a safe haven to retreat to when alarmed — and the system is goal-corrected, continually adjusting the child's behavior to keep a comfortable degree of closeness given the level of threat. Attachment behaviors surface most clearly when the system is activated by fear, fatigue, illness, or separation, which is precisely why a laboratory procedure built around brief separations would prove so revealing.
Attachment · the missing styleThere are four styles, not three
To turn Bowlby's theory into something measurable, Mary Ainsworth devised the Strange Situation, a twenty-minute laboratory procedure of eight short episodes that gently escalates a baby's stress through two brief separations from the caregiver and encounters with a stranger in an unfamiliar room (Ainsworth et al., 1978). The logic is that mild stress activates the attachment system, and the telling moment is not the separation itself but the reunion — how the one-year-old uses, or fails to use, the returning parent to recover. Ainsworth's coding of North American infants yielded three organized patterns, each a coherent strategy for managing the caregiver's availability.
The first and most common is secure attachment: the baby uses the parent as a secure base to explore, shows genuine distress at separation, and — the defining feature — is readily comforted at reunion before returning to play. In Ainsworth's earlier home observations this pattern was predicted by maternal sensitivity: caregivers who noticed their infant's signals and responded promptly and appropriately tended to have securely attached babies. The two insecure but organized patterns are strategies for coping with less reliable care. Insecure–avoidant infants show little overt distress and snub or ignore the parent at reunion, a down-regulation of attachment needs associated with caregiving that is consistently distant or rebuffing. Insecure–resistant (or ambivalent) infants are intensely distressed and cannot settle at reunion, seeking contact and angrily resisting it at once, a pattern linked to inconsistent, unpredictable responsiveness.
For years these three categories were the whole story, but a stubborn minority of infants could not be forced into any of them — they would freeze, approach the parent and then veer away, rock, or adopt a dazed, trance-like expression. Reanalyzing hundreds of these "unclassifiable" tapes, Main and Solomon added a fourth category, disorganized/disoriented attachment, defined by the momentary collapse of any coherent strategy in the parent's presence (Main & Solomon, 1990).
Mary Main and Judith Solomon added a fourth category: disorganized/disoriented attachment. These infants show contradictory, fearful, or dazed behavior toward the caregiver — the person who should be the safe haven is also the source of alarm, so the attachment system has no coherent strategy. It's the style most strongly linked to caregiver trauma/frightening behavior and to later psychopathology. When you see "three styles," that's the pre-1990 picture.
The four attachment styles in the Strange Situation. Tap one to see the behavior pattern.
The classic finding that ~65% of babies are secure comes mostly from Western, middle-class samples. Van IJzendoorn & Kroonenberg's (1988) landmark meta-analysis found the distribution shifts by culture: German samples showed more avoidant classifications (independence is prized early), while Japanese and Israeli-kibbutz samples showed more resistant classifications (babies rarely left alone with strangers, so separation was more alarming). Attachment security itself is a robust human phenomenon — but the Strange Situation is partly reading cultural values about separation, not just relationship quality.
Attachment · what it predictsDoes the Strange Situation matter?
A laboratory classification is only interesting if it forecasts something, and attachment security does — modestly but reliably. Disorganized attachment in particular has drawn attention because it is the pattern most consistently tied to adverse caregiving and to later difficulty: a large meta-analysis found it far more common where parents are maltreating, depressed, or behave in frightening ways, and identified it as the infancy classification most predictive of later externalizing problems and psychopathology (van IJzendoorn et al., 1999). Security, by contrast, forecasts on average better peer relationships, emotion regulation, and social competence in childhood, though the effect sizes are moderate and no single early classification determines a life.
Two cautions keep these predictions honest. First, attachment is a property of a relationship, not a fixed trait stamped on the child: a baby can be securely attached to one parent and insecurely attached to the other, which is hard to reconcile with any purely temperamental account. Second, the cross-cultural data warn against reading the Strange Situation as a culture-free thermometer of relationship quality. Because the procedure measures reactions to separation, and cultures differ in how much solo separation infants normally experience, the same reunion behavior can carry different meaning in a German, a Japanese, or a kibbutz sample — which is why the distribution of styles, while never lacking a secure majority within a sample, shifts systematically across them (van IJzendoorn & Kroonenberg, 1988).
The bridge forwardInternal working models & adult attachment
The reason attachment reaches far beyond infancy is a cognitive one. Out of thousands of repeated interactions, Bowlby argued, the infant abstracts an internal working model — a largely unconscious mental representation carrying two linked expectations: whether others are likely to be available and responsive, and whether the self is the kind of person worth responding to (Bowlby, 1969). These models function as templates, shaping what a person expects from new relationships, how they read a partner's ambiguous behavior, and how they regulate closeness and distress. Because they operate outside awareness and tend to be self-confirming — an anxious model elicits reassurance-seeking that can strain the very bonds it fears losing — working models persist, giving early attachment its long reach into later life.
Two ways to measure adult attachment
Adult Attachment Interview (AAI) — a structured interview scored not on what happened in childhood but on how coherently you can tell the story. Predicts how you'll parent your own kids.
ECR self-report — questionnaires (e.g., Experiences in Close Relationships) that place adults on two dimensions: anxiety and avoidance, rather than in boxes.
The best news in psychology
Attachment style is moderately stable, not fixed. Longitudinal studies show style can change with major life events, therapy, or a stable loving relationship — the "earned secure" pattern, where someone with a rough childhood develops a coherent, secure stance as an adult. Early ≠ destiny. That's the idea worth holding onto.
The conceptual bridge to adulthood was built when romantic love itself was reconceived as an attachment process. Adults, on this account, relate to romantic partners much as infants relate to caregivers — seeking proximity, using the partner as a secure base and a safe haven, and protesting separation — and the three-way typology of secure, avoidant, and anxious reappears in how adults describe their close relationships, in roughly comparable proportions (Hazan & Shaver, 1987). That parallel is why attachment is the through-line connecting this infancy unit to the emerging-adulthood and relationships units later in the course: the working model built with a first caregiver becomes, revised and elaborated, the lens through which a person approaches partnership, marriage, and eventually their own children.
None of this makes early attachment a life sentence, which is the single most important thing to take from the topic. Longitudinal work shows that attachment is moderately stable rather than fixed: models formed early bias but do not dictate later relationships, and change is not only possible but common, tracking major life events such as loss, a stable and loving partnership, or psychotherapy (Fraley, 2002). The clearest evidence is the earned-secure pattern — adults who describe a difficult, insecure childhood yet, having reflected on it, can now tell a coherent and balanced life story and relate securely in the present. The Adult Attachment Interview (George et al., 1985) scores exactly this coherence rather than the pleasantness of the history it recounts, and coherence, not a happy childhood, is what best predicts whether a parent will raise a securely attached child of their own. Early experience matters, but it is a first draft, not the final text.
Harlow's wire-vs-cloth-mother monkeys gave us "contact comfort" and helped end the sterile, no-touching model of institutional childcare — a genuinely important legacy. But it's worth naming that the maternal-deprivation and "pit of despair" studies inflicted severe, lasting harm on primates and are frequently cited as a turning point in research-ethics reform. Good science literacy means holding the finding and the ethical cost in view together.
SourcesCited in APA 7
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the Strange Situation. Erlbaum.
Berko, J. (1958). The child's learning of English morphology. Word, 14(2–3), 150–177.
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.
Chomsky, N. (1959). A review of B. F. Skinner's Verbal Behavior. Language, 35(1), 26–58.
DeCasper, A. J., & Fifer, W. P. (1980). Of human bonding: Newborns prefer their mothers' voices. Science, 208(4448), 1174–1176.
Eimas, P. D., Siqueland, E. R., Jusczyk, P., & Vigorito, J. (1971). Speech perception in infants. Science, 171(3968), 303–306.
Fraley, R. C. (2002). Attachment stability from infancy to adulthood: Meta-analysis and dynamic modeling of developmental mechanisms. Personality and Social Psychology Review, 6(2), 123–151.
George, C., Kaplan, N., & Main, M. (1985). The Adult Attachment Interview [Unpublished manuscript]. University of California, Berkeley.
Harlow, H. F. (1958). The nature of love. American Psychologist, 13(12), 673–685.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524.
Kuhl, P. K. (2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5(11), 831–843.
Kuhl, P. K., Williams, K. A., Lacerda, F., Stevens, K. N., & Lindblom, B. (1992). Linguistic experience alters phonetic perception in infants by 6 months of age. Science, 255(5044), 606–608.
Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years (pp. 121–160). University of Chicago Press.
Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical learning by 8-month-old infants. Science, 274(5294), 1926–1928.
Skinner, B. F. (1957). Verbal behavior. Appleton-Century-Crofts.
Tomasello, M. (2003). Constructing a language: A usage-based theory of language acquisition. Harvard University Press.
van IJzendoorn, M. H., & Kroonenberg, P. M. (1988). Cross-cultural patterns of attachment. Child Development, 59(1), 147–156.
van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999). Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants, and sequelae. Development and Psychopathology, 11(2), 225–249.
Werker, J. F., & Tees, R. C. (1984). Cross-language speech perception: Evidence for perceptual reorganization during the first year of life. Infant Behavior and Development, 7(1), 49–63.