Loss of inflections
The noun system of Old English was quite complex with 3 genders (masculine, feminine and neuter) and 5 cases (nominative, accusative, genitive, dative, instrumental). In the history of English this was simplified considerably. The cases were reduced to nominative and genitive and the phenomenon of grammatical gender was lost.
In addition to gender and case Old English had a variety of plural types just like Modern German. The number of these has steadily declined throughout the centuries. This demise can be documented quite clearly and the reduction in diversity shows a definite sequence which can be summarised as follows:
man ~ men
child ~ childer
ox ~ oxen
stone ~ stones
It is clear from Old English that the umlaut plurals declined quite rapidly. We have words like cu with an original plural cy (compare German Kuh ~ Kühe) which later becomes cyne (cf. the Early Modern English form kine). /r/ plurals are replaced by nasal plurals in the early Middle English period as the present-day form children shows which has a nasal ending added to an original childer (/r/-plural). The nasal plurals themselves pass into decline by the late Middle English period (Chaucer still has eyen for modern eyes).
Now note that the umlaut plurals are an inherited type from pre-Old English forms of Germanic which English shares with other forms of North and West Germanic like German and Swedish. From the point of view of iconicity (a form indicates a grammatical category) the umlaut plurals fulfill their function well. The singular and the plural forms of words are clearly distinguished.
However, from the point of view of the language’s grammar the umlaut type contravenes a principle of morphology which requires that there be an isomorphic relationship (one to one in form) between the lexical root or stem of a word and any inflection added to it. Ideally, this type of situation applies to agglutinative languages. Looking at the remaining three formational types one sees that they form a scale of decreasing sonority (open, vowel-like quality): /r/ -> /n/ -> /s/. A general principle of morphology is that inflectional endings are favoured which show a high degree of phonetic salience. It is this factor which increases with progression in types just indicated.
The point being made here is that any explanation which tries to avail of either phonological or morphological arguments exclusively is doomed to failure as there was a morphological reason for loss in the first case and a phonological reason in each of the remaining three.
The effects of the above changes on the morphology of Middle English were very considerable. They led to a loss of distinctiveness among grammatical endings so that the various declensional classes of Old English collapsed, with the dative plural remaining for a while the only case — with a final nasal /-n/ — which was distinctive, but even that was reduced in the course of the Middle English period. A direct consequence of this was that the more common declensions were generalised and used productively. The two main ones are the s-type and the nasal type as seen in the Old English words stān ‘stone’ : stānas ‘stones’ and ēage ‘eye’ : ēagan ‘eyes’ respectively. For a while the nasal declension was productive as is seen in its addition to the old r-plural child : childer > child(e)ren to give the doubly marked plural which has survived to the present day. The north of the country was as always innovative and by about 1200 nouns are commonly found with a plural and a genitive singular in /-s/, this then spreading to the south somewhat later and with time it replaced virtually all nasal plurals. There are a few remainders into the time of Shakespeare — e.g. eyen, shoon, housen — but these have been brought into line with the universal s-plural so that nowadays there are only three nasal plurals remaining: oxen, children and brethren (a double plural with an umlaut of brother and a nasal ending).
There are a few other plural types of which reflexes still exist in English. Most noticeable are umlaut plurals which are the forerunners to the modern word pairs foot : feet, goose : geese, man : men, mouse : mice. These nouns are part of the core of English vocabulary and are nearly all terms for humans, parts of the body or familiar animals. Still less significant are the few examples of zero plurals, all terms for animals as in sheep : sheep, deer : deer, fish : fish (an analogical plural form fishes also exists).
Be careful to distinguish these instances of inherited plural types from cases which are derived from direct imports from Latin or Greek. Hence in Modern English one has pairs like formula : formulae, criterion : criteria which are direct loans from Latin and Greek respectively and show the plural endings typical of these languages.
The reduction in morphological variation which is found with nouns applied to other word-classes as well. Adjectives lost their endings so that the previous distinction between a strong and weak declension — as with Modern German dichter Nebel and der dichte Nebel — was lost.
Equally one can notice a loss in grammatical gender in the transition from Old to Middle English. The older stage of the language showed three genders as in Modern German, masculine, feminine and neuter, distributed on arbitrary grounds, e.g. the word wīf was neuter (cf. German das Weib). There were three forms of the definite article þe, sē and sēo. By the end of the Middle English period there was only one form, the modern the (which derives from Old English þe). The consequence of the loss of grammatical gender is that it was replaced by natural gender in most instances. There are examples in Modern English in which another gender is used — for instance, a feminine reference is used for technical objects such as cars, planes or ships — but this is more an analogical extension of natural gender rather than a survival of grammatical gender.
ALLOPHONES TO PHONEMES Word-medial fricatives [v, z, ð] were allophones in Old English, e.g. līf ‘life’ ended in [-f] whereas libban — cf. the preterite form with the voiced fricative lifde — had an internal voiced fricative [-v-]. However with the loss of the inflectional endings the verb was reduced to a single syllable and the final /-v/ now contrasted with the final /-f/ of the noun, hence the change in status of voice among fricatives in Middle English which from then on distinguished phonemes.
The system status of voiced and voiceless fricatives was strengthened by the fact that with French loans instances of voiced initial fricatives now occurred in English, adding to the functional load — and hence to the systematic importance — of these segments.
vertu ‘virtue’ vileynye ‘villainy’ zēle ‘zeal’
Another development relevant to this issue is the loss of distinctive consonant length in Middle English. Recall that in Old English consonants could be long or short, e.g. /-s-/ [-z-] and /-ss-/ [-ss-] were phonemically distinct. cf. Offa (proper name), missan ‘miss’, siþþan ‘sit’ which show geminates (long consonants) for all three fricative types in Old English. In the Middle English period this distinction begins to be lost so that the instances of voiceless word-medial geminates were reduced to simple segments but remained voiceless, so that the original phonetic contrast between voiced medial non-geminate versus voiceless medial geminate was now reduced to a simple distinction between voiced and voiceless fricative in medial position, thus strengthening the phonemic significance of voice for these fricatives.
Later initial /θ-/ in grammatical formatives such as the, there, that, etc. was softened to [ð] because of the unstressed character of the words. Thus the series /v-, ð-, z-/ in initial position was completed.
THE WORD FOR ‘WOMAN’ This derives from Old English wīfmann ‘woman’ + ‘man’ (German ‘Weib’ + ‘Mann’). The plural was formed by umlaut on the second element, i.e. wīfmenn (cf. German Mann : Männer). Now in the course of time English lost the distinction between /a/ and /e/ in unstressed syllables — both collapsing to schwa /ə/ — and in addition in this form the fricative /f/ was lost before the nasal /m/ and the vowel of the first syllable was shortened, so that the form was then /wɪmən/ and homophonous in the singular and plural in many varieties of English. Now the initial /w/ had a retracting influence on the first vowel so that a form arose with a high back vowel rather than a high front vowel, i.e. /wʊmən/. This came to be used for the singular of the noun and the original pronunciation for the plural form, so that one had /wʊmən/-SG versus /wɪmən/-PL. With the lowering of /ʊ/ in the 17th century, most words adopted a value /ʌ/ for the former high vowel. However, the pronunciation of the singular of woman was retained as /wʌmən/ (probably under the influence of the preceding /w/, cf. would /wʊd/). Note that the spelling in modern English, woman : women, implies that the plural has a change in the vowel of the second syllable. This is not the case, the vowel of the first syllable changes, the spelling is just to form a parallel orthographical case to man : men (here the phonetic alternation is indeed between /æ/ and /e/ for the singular and plural respectively).
Compensation for the loss of inflections
If grammatical categories were indicated in Old English via inflections then the loss of the latter implied that something took their place. The answer to the question ‘what?’ is simple: word order and the increased functionalisation of prepositions. In Old English the order S - O - V (Subject - Object - Verb) was common but with the loss of inflections the indentification of Subject and Object was not always that simple. For this and other contributory reasons the order S - V - O (Subject - Verb - Object) became more usual in the course of the Middle English period. The order V - S - O (Verb - Subject - Object) which was also found in Old English declined in frequency, remaining most tenaciously after adverbs where it is still sometimes found today as in Hardly had he left the room when she rang.
The increased use of prepositions served the function of rendering sentences unambiguous. A simple example illustrates this. The German sentence Er schrieb ihr einen Brief has variants like Ihr schrieb er einen Brief or Einen Brief hat er ihr geschrieben all of which are possible because the inflected forms of the pronouns and the object noun are unambiguous with regard to sentence function. In English there are two equivalents to the sentence He wrote her a letter and He wrote a letter to her, the former uses word order to indentify the sentence elements functionally — indirect object precedes direct object — and the latter employs a preposition to to identify the indirect object. For topicalisation as in the German examples modern English has to resort to intonational strategies (stressing the highlighted element) or to the syntactic device of clefting which retains the prepositional object but moves it to the front by embedding it into a dummy sentence: It's to her that he wrote the letter.