There are numerous rankings of “difficult” (and “easy”) languages to learn (note that we are talking about second-language learning here, not acquiring one’s native tongue). Some such rankings are unofficial, like the Accreditedonlinecolleges.com ranking or the ranking at MyLanguages.org; others are official, for example the classification by DLI based on the number of hours needed to achieve a certain level.
The top spot in all these ranking is given to Chinese: Accreditedonlinecolleges.com lists Mandarin and Cantonese separately, as #1 and #2; MyLanguages.org lists Chinese at the top spot and the DLI classification includes un-specified “Chinese” in the most difficult group IV, together with Arabic, Japanese and Korean. But what are the main perceived difficulties in learning Chinese (whether Mandarin, Cantonese or another Chinese variety)? In addition to the general themes of an unfamiliar writing system and dialectal variation (expected of a language with about a billion speakers, like Mandarin, or even “smaller” Chinese languages: Cantonese with 52 million speakers; Shanghainese with over 77 million speakers; Taiwanese with over 25 million speakers), Chinese is difficult to learn because of the “exotic” sounds it has and especially its tone system. There are four tones in Mandarin and six tones in Cantonese.
What about learning other East Asian languages, like Japanese or Korean? These too are typically listed at the top of the difficulty ranking. Among the most difficult aspects of Japanese (in addition to its three-part writing system, including kanji, hiragana and katakana) are “an agglutinative vocabulary” and “rigid hierarchical structure of honorifics inextricably tied to Japanese society and culture”. I agree that the rich system of honorifics — markers of esteem or respect when used in addressing or referring to a person — can be difficult to learn because of the cultural knowledge that one needs in order to be able to use these grammatical forms correctly. You will run into this difficulty if you attempt to learn Japanese, Korean or Thai.
But I am not quite sure what the problem is supposed to be about the “agglutinative vocabulary”. In fact, the term “agglutinative” typically refers not to vocabulary but to the morphological system of a language. In an agglutinative language, each affix attached to the root typically expresses one grammatical property, such as gender, number or case, but not all three at once (as would be the case in a fusional language; see below). Multiple affixes can be attached to the root, and when it happens, affixes do not have much effect on each other’s pronunciation or meaning. Other examples of languages that are agglutinative and are thus said to be difficult to learn include Basque, Korean and Hungarian.
Some, though not all, agglutinative languages also have rich systems of case markers, that are used not only to mark such grammatical functions as the subject, object, indirect object, possessor, etc. but also to indicate spatial relations. While some “difficulty rankings” will scare you with statements like “Basque’s complexities … lay in its 24 cases” or “anyone hoping to pick up Hungarian must also completely conquer its whopping 35 cases”, some of these statements overexagerate. The consensus among linguists is that Basque has only 12 cases and Hungarian has only 21 cases. Two of Hungarian’s cousins — Finnish and Estonian — have somewhat “poorer” cases systems (14 cases in Estonian, 15 cases in Finnish), while others have even richer case systems: for example, certain dialects of Komi have up to 27 cases. By the record-holder in terms of the number of cases is a Dagestanian language Tabasaran with its 46 cases!
While it may appear daunting to learn so many cases, in Finno-Ugric languages, such as Finnish or Hungarian, the form of the case morpheme is the same regardless of what noun it attaches to. The situation is quite different in Slavic languages like Russian or Polish, which have fewer cases (for instance, Russian has “mere” six cases), but the forms of the case morphemes differ depending on what noun you attach the case morpheme to (these different types of nouns are known as “declension patterns”, and they are closely related to but not exactly the same as “genders”). For example, the dative (singular) in Russian is -e if it attaches to knig- ‘book’, -u if it attaches to stol- ‘table’ or -i if it attaches to mater- ‘mother’. Thus, instead of learning one set of 15 case affixes in Finnish, for Russian you need to learn three sets of six (i.e., 18) case affixes, as well as to know which affixes to attach to which nouns. Besides, to master the Russian case system you will need to learn the various exceptions, which too are more common in fusional languages like Russian than in agglutinative languages like Finnish. My conclusion: agglutinative languages may be somewhat easier to learn (at least, in terms of the memory load) than fusional languages; the only truly scary thing about agglutinative languages is the term itself!
And what of Arabic? One of chief reasons it lands in one of the top spots in the “difficulty ranking” is the script, which uses different shapes of letters word-initially, -medially and -finally, but has no letters to record vowels. Another difficulty often listed is the dialectal problem; however, most students of Arabic as a foreign language will be learning Modern Standard Arabic rather than one of the 40 or so colloquial varieties, spoken from Morocco to Egypt, from Syria to Iraq and the Persian/Arabic Gulf. Grammatical difficulties one should be prepared to face include the unfamiliar Verb-Subject-Object order (vs. the English Subject-Verb-Object order), dual number (in addition to the familiar singular and plural), three cases and two genders (which, all in all, should be much easier to learn than the Hungarian or Russian case+gender systems) and multiple verbal forms. A really unusual phenomenon that Arabic shares with Hebrew is its non-concatenative morphology. In a non-concatenative language, unlike in more familiar languages such as English, Spanish or Russian, grammatical meaning -– for example, plural number on nouns or past tense on verbs -– is expressed not through adding a suffix to the nominal or verbal root/stem, but through changing the vowels in the stem. Typically, the consonants are part of the noun or verb root, while vowels -– and where they are placed in relation to the consonants of the root -– constitute the “template” (hence, non-concatenative morphology is also known as “root-and-template morphology”).
Finally, one other language that I found in one of those lists of “difficult to learn languages” — to my great surprise — was Icelandic. It is certainly true that “many Icelandic phonemes don’t have exact English equivalents” — remember the difficulties that many American journalists had with that Eyjafjallajökull volcano?! Other perceived difficulties of Icelandic include “its archaic vocabulary” and “complex grammar”. Really? It is true that Icelandic is one of the most conservative North Germanic (i.e., Scandinavian) languages that has kept the old noun declension and verb conjugations, but it is undoubtedly much closer to English and much more similar to it than, say, Chinese, Arabic or Hungarian, which makes it rather easier for an English speaker to learn.